407: It Isn't a Big Grapefruit


00:00:00   My iMac fan noise problem is worsening.

00:00:03   Oh, so I guess you're buying a new Mac then. Probably a Mac Mini?

00:00:06   No, he just needs to take it off of his desk, put it into a case and carry it around on the street for a while and then unpack it.

00:00:11   That's what fixed it last time, right?

00:00:13   Actually, yeah. The transportation to the beach did, I guess, dislodge enough dust.

00:00:18   I don't know, I'm getting near the end of my rope here. It's pretty bad. It's now audible most of the day, even when it's ostensibly idle.

00:00:28   Like, when there's nothing in activity monitor or iStack menu showing it's actually using the CPU, it's now idling at like 72 degrees Celsius and like 13 to 1400 RPM fan speed, which is not normal.

00:00:43   And this isn't even just Catalina being Catalina because it persists through reboots too.

00:00:48   So I'm just, this is the worst time in possibly a decade to need a new high performance desktop Mac.

00:00:59   Yeah.

00:01:00   I did actually, because I actually looked, the trade-in value for my iMac Pro is almost $3000.

00:01:09   Wow, that's not bad actually.

00:01:11   I know. I was like, that's actually, I could just trade it in and just buy like, you know, the highest spec Mac Mini with the new chip, which is like $1700.

00:01:22   But I would only then have two terabytes instead of four and I would have only 16 gigs of RAM instead of 64.

00:01:28   And I'm a little hesitant to go that route. And I'd have to, you know, go get my stupid LG 5K and bring it to the beach.

00:01:36   No way in hell I'm ordering a second one, but you know, just move that one here.

00:01:41   So it's, it would be kind of an ordeal. And then like what I really want ideally is a new iMac.

00:01:50   Like I love the iMac, although, and well, do I love the iMac?

00:01:55   I mean, I've had now two, I guess, right? Did I only have one 5K before the Pro?

00:02:02   No, I thought you had two.

00:02:03   You had the one with image retention and I don't remember if you ever got that fixed.

00:02:06   I did eventually get it fixed. That one came out in, what was it? 2014 the first 5K came out? Something like that?

00:02:13   Yeah.

00:02:14   The iMac Pro came out in the end of 2017 and I had it for three years. So yeah, so I'm pretty sure I just went straight from that iMac 5K, the first gen iMac 5K to the iMac Pro.

00:02:23   And in both cases I loved them for a while, but they didn't age incredibly gracefully because of the all in oneness. Like, you know, the first one I had, I had a screen problem and it's, I, you know, it took a while for me to be able to go get a fix.

00:02:38   I don't want to send my whole iMac in for repair for, you know, just the screen. Like the computer part was fine.

00:02:44   If it was a Mac Pro, I could just unplug the monitor, you know, like the all in oneness there kind of limited its lifespan.

00:02:50   And in this case, the all in oneness is limiting its lifespan as well. So maybe I should learn from this. I probably won't.

00:02:55   What I ultimately like, if there was an M1 based iMac, I would have probably already ordered it.

00:03:02   As we, as we record, and I've mentioned this before, I try to work on the show notes and get a pretty solid rough draft ready.

00:03:08   And I had already written, you heard it here first, Marco is really starting to justify a new Mac.

00:03:14   I think I will now change that to you heard it here first. Marco is really starting to justify a new Mac Pro.

00:03:20   I will not buy a Mac Pro. I would buy the 6K HDR display before I buy the Mac Pro, which is funny because if I did that,

00:03:30   what I would plug in the 6K HDR display with its fan to would be a MacBook Air that I have that has no fan.

00:03:40   My monitor would have a fan on my computer. Wouldn't it?

00:03:43   You can't hear the monitor fans. Don't worry. I'm still telling you. I'm sitting in front of it right now.

00:03:49   I was looking up iFixit today. It's like, how do you open an iMac Pro if like, I wanted to get in there and actually try to remove like, you know, major dust buildup.

00:03:58   How do I get in there and how risky is it? And the the answer is like, it's a pain in the butt and moderately.

00:04:06   And so I don't I don't know if I want to attempt that, especially in a dusty, salty air environment that's highly corrosive in a house that I have very few tools and no abilities to open iMacs, you know, effectively and cleanly.

00:04:21   It's probably a bad idea. So I'm probably not going to want to attempt that myself. So the the end of the world, notwithstanding, would it be cheaper to fly a Stephen Hackett to you to perform this operation and then fly him home?

00:04:35   I'm just let's think a little outside the box.

00:04:39   Well, it would probably be more practical to just, you know, go take my iMac, put it in the carrying case, walk it to the ferry, get on the boat, go across the bay, go to the nearest Apple store, drop it off or I guess I could do mail in.

00:04:52   But that's that's a whole other ordeal. And then, yeah, and have Apple do it. I'd be without it for a while.

00:04:57   Then I'd have to take another set of, you know, ferry and car trips and walking trips and to go pick it up in a few days whenever it was done. So it would be an ordeal.

00:05:08   I'd rather not do it. I mean, really, the most responsible thing to do here would just be to just deal with my stupid fan noise for the next whatever.

00:05:16   It'll be seven months, maybe before there's going to be an iMac pro or an iMac with the M series chips in them.

00:05:23   But man, this this is annoying. I'm sorry, but I'm quite sure that the moment something that even vaguely smells like an iMac with Apple Silicon in it is released.

00:05:35   I am sure that you will be the very first person in line to replace your your poor struggling iMac pro.

00:05:42   I mean, I really am. I have been very tempted to just switch to my air full time somehow, even though it only has one terabyte. And, you know, I could use externals, I guess I could I could make that work with some hacks and then just somehow solve the monitor situation.

00:06:00   And which would probably be on my next trip, picking up my my five can bring it here. So maybe that's what I'll do. But I don't know. I don't love that option.

00:06:09   What do you hate more? Do you hate fan noise more? Or do you hate that LG 5k?

00:06:13   That's a good question. I, I really don't love the idea of buying the 6k monitor to have an unknown future computer. Like what am I going to plug into that in a year?

00:06:25   What if the most compelling computer for me in the next year ends up being an iMac again, then I bought this giant monitor with all this money for, you know, only six months of use. And then like that's that's a pretty big loss and a pretty big waste.

00:06:40   I think like I would have trouble swallowing that. Whereas if I just either had a Mac Mini that I could plug in or something like that or plugged in my MacBook Air to the LG 5k. I don't like swallowing something about all these plans.

00:06:56   But I think using all of the hardware I already have that is like the LG 5k just sitting in my house like in Studio B doing nothing. I could just use that. And that's that's probably what I should do.

00:07:10   Even though like I'll be grumbling the whole time looking at stupid monitor and it's terrible backlight leakage and it's giant ugly frame and the big ugly foot and all that stuff.

00:07:19   So, Jon, would you mind going to apple.com/shop/trade-in? I'm curious if you enter in, Jon, the serial number of your XDR. I doubt it will like magically make it through. And please do not, you know, share that obviously verbally.

00:07:34   But if you just quietly enter in your serial number, I wonder what it would say.

00:07:38   Actually, it does have an option for it. Under other devices, it says what kind of device do you want to use? You say displays and it says recycle.

00:07:47   Right, exactly. That's recycling only, which is not what you want. You can recycle your display, Jon.

00:07:51   Which model do you I entered my serial number and it says which model do you have? And now I have to pick other? Yeah, I guess.

00:07:57   Based on what you've told us today, your Mac is ready to recycle.

00:08:01   I entered a display serial number, so I don't know what you're talking about. That's funny. All right. Well, it was worth a shot.

00:08:08   All right. Shall we start with some follow up? We have hold. We have follow up today. It'll be a miracle if we get out of follow up today.

00:08:14   So let's start. Ryan Fegley writes, I had some duplicate, the same duplicate contacts issue as Marco this week caused by an upgrade to Big Sur.

00:08:22   I believe deduping the contacts or deduping in contacts didn't work.

00:08:27   So I restored from a pre upgrade iCloud backup on the iCloud Web site in a browser. I had no idea this was possible.

00:08:35   Ryan writes, it worked like a charm. It's good to know it's an option. So you log into iCloud.com.

00:08:39   You click on your name in the upper right and then go to account settings. We'll put a link in the show notes.

00:08:44   And then it says restore blank on the bottom left. Restore files, restore contacts, restore calendars, restore bookmarks.

00:08:50   And apparently that's a way you can restore stuff, which is news to me. I had no idea.

00:08:55   I filled in the details this because I'm like, I remember being on iCloud.com, being able to restore stuff.

00:09:00   It was it was a little bit of a trick to find like you wouldn't think to go to account settings.

00:09:05   And then then these these links appear totally on the opposite side of the page, but they do.

00:09:09   So you can restore. I think you can restore individual files. So, you know, like it's nice that this stuff is there.

00:09:14   I don't understand why it's not on the Mac, but whatever. Second bit of follow up related to this is that when I was doing this to confirm it,

00:09:21   I opened up contacts. And guess what? I had duplicates of all my contacts, too. I'm like, wait a second.

00:09:25   I didn't even upgrade. I didn't even upgrade to Big Sur. Why the hell do I have duplicates?

00:09:29   And I said, oh, yeah, I didn't upgrade to Big Sur, but I just got an M1 Mac that runs Big Sur.

00:09:35   And I made my account on it and I logged in and connected it to my iCloud. So something having to do with Big Sur is really, really excited about making duplicate contacts.

00:09:45   I just used the Ddupe feature in contacts and it fixed my problem. I have no idea why I had duplicates, but that's what I had.

00:09:51   Cool. Lucky us. I don't seem to have that problem. Yeah. Lucky you guys.

00:09:55   Well, do you have you, Casey, have you logged into a Big Sur, an official release Big Sur machine with your actual Apple ID yet?

00:10:03   Yeah, because my MacBook Pro has been on Big Sur for like a week, maybe two.

00:10:07   All right. Well, you like that. But me and Marco drew the duplicate straw.

00:10:10   So sorry. All right. So accelerated TensorFlow for Mac OS probably just uses the GPU.

00:10:16   If I recall correctly, we were surmising that maybe it uses the neural engine.

00:10:20   And apparently that's not the case. There's a link to machine learning dot Apple dot com where they write until now, TensorFlow is only used to utilize the CPU for training on the Mac.

00:10:29   The new TensorFlow underscore Mac OS fork of TensorFlow two point four leverages M.L. compute to enable machine learning libraries to take full advantage of not only the CPU, but also the GPU in both M1 and Intel powered Macs for dramatically faster training performance.

00:10:44   Very cool. Yeah. And some people were saying on Twitter like, oh, it's because the neural engine can't be used for training.

00:10:49   But his Nash on Twitter says neural engine can be used for training as well, but it's limited in the type of operations.

00:10:53   So M.L. compute framework uses the CPU and GPU because it's more generic.

00:10:58   So, yeah, TensorFlow did get a lot faster and it is seven times faster on the M1 Mac, but it's not because of the neural engine.

00:11:04   Indeed. Max Lean writes that Marco said on this is actually this is a very good well, actually, this is the kind of well, actually, I can get behind.

00:11:12   Marco said in the podcast that Apple has never made memory controllers with support for ECC RAM.

00:11:16   Admittedly, it has been a while, but the Power Mac G5 and Xserve G5 both supported ECC RAM and the memory controller was made by Apple.

00:11:22   Yeah, I should have thought of the power. When he said that, I was trying to pull something out of my head of whether they did.

00:11:27   And I remember the Power Mac G5 had ECC RAM, but I couldn't remember if Apple made the controller, if it was some Motorola or IBM thing.

00:11:32   But we'll trust Max. He seems like he knows.

00:11:35   We are sponsored this week by Linode, my favorite place to run servers.

00:11:40   Whether you're working on a personal project or managing enterprise infrastructure, you deserve simple, affordable, and accessible cloud computing solutions that allow you to take your project to the next level.

00:11:50   Simplify your cloud infrastructure with Linode's Linux virtual machines to develop, deploy, and scale your modern apps faster and easier.

00:11:57   You can get started today on Linode with $100 in free credit at linode.com/atp or by texting ATP to 474747.

00:12:10   And that gets you instant access to $100 in free credit.

00:12:14   Now, I've been with Linode for a long time. $100 goes a long way because their plans start at just $5 a month.

00:12:20   And they have all sorts of amazing specialty capabilities if you need those or just straight up, you know, some RAM, some disks, some CPU.

00:12:28   Whatever you need, Linode has you covered. They are by far my favorite web host I have ever used.

00:12:34   And I've used a lot of web hosts. But I run a lot of servers and I choose to run all of them on Linode now because it's just that good of a host.

00:12:43   They have 11 data centers worldwide, 24/7, 365 human support, and their support team is excellent.

00:12:50   I've used them, you know, over the years. I've been with them almost a decade. So I've filed a few support tickets here and there.

00:12:55   And they always have incredibly knowledgeable people. There's no like tiers or levels you've got to like, you know, work your way through.

00:13:01   It's always top, top, top notch support right from the very first person who responds. And they respond quickly and they are really great.

00:13:07   See for yourself once again, Linode.com/ATP and click the create free account button to get started. You can get $100 in credit.

00:13:17   If you aren't able to write this down, just try to remember text ATP to 474747. Get started on Linode today.

00:13:26   Thank you so much to Linode for sponsoring our show and hosting all of my servers.

00:13:34   All right, moving on. Jonathan Dietz writes, "After listening to the scaling the Mac, the M1 architecture to higher end machine section from episode 406, I figured I'd send along some cheat sheets of relevant info.

00:13:44   Die size is generally constrained by the reticle limit, which is due to current lithography steppers having a maximum field size of 26 millimeters by 33 millimeters or 858 square millimeters.

00:13:55   It's important to keep in mind just how many different chips Apple is integrating into each M series system on a chip. The CPU, the GPU, the PCH, the T2. What's a PCH?

00:14:04   Platform Controller Hub, I think. It's like what used to be called the Southbridge, I think.

00:14:09   Ah, right. Okay, thank you. T2 functions including storage controller and the Thunderbolt/USB4 controllers.

00:14:14   It's also worth noting that decent SSD controllers like the one in the T2 have their own memory interface for DRAM cache.

00:14:21   With the M1, the storage controller is connected directly to the system fabric and shares the unified memory pool, meaning performance is no longer limited by PCIe x4 connection to the host processor or cache size.

00:14:31   Eliminating the need for both integrated and discrete GPUs, as well as consolidating the two or three different memory subsystems has a lot of upside for Apple.

00:14:39   There's a spreadsheet that we'll link in the show notes. There's a fancy chart that, if we remember, Marco will make the show art for this chapter.

00:14:46   With regard to RAM, good old DDR4 DIMMs are the only way to scale capacity. Apple will likely go with four HBM2E stacks. What in the world does that mean?

00:14:56   A couple of people were talking about HBM. Well, we'll talk more about that in the next bit here.

00:15:02   High bandwidth memory. It's not a great acronym because what does that even mean? What does high mean?

00:15:08   What happens when the next memory interface comes out and it has higher bandwidth than supposed HBM or whatever?

00:15:14   You call it ultra high bandwidth 4K. Super speed, yeah, we know.

00:15:19   So, anyway, Apple will likely go with four HBM2E stacks for up to 64 gigs of crazy fast on-package memory with the M1X, but would need to tack on an additional eight channel DDR4 interface for any hope of parity with the Intel Xeon or AMD E-PYC. Is that supposed to be Epic platforms?

00:15:37   That's how I pronounce it when I read it.

00:15:39   It's also worth noting that regardless of memory type, the highest density DRAM dies currently available are only 16 gigabit or two gigabytes. Increasing capacity means adding and paying for more dies. There's no way around that.

00:15:50   So I did actually make the image that's in our show notes here is perfectly square for the purposes of show art. So, Marco, please make this graph.

00:15:58   Because I spent so long making this stupid graph because I do not know how to use spreadsheets to graph things and they make me so angry.

00:16:04   I always go to numbers because I see Jason Stone makes his graphs in numbers. I'm like, "Oh, Apple's nice. Everything's going to look nice. I'll use numbers to make it." But I cannot figure out how to do anything in numbers.

00:16:15   So then I go to Excel grudgingly, and I also don't know how to use Excel at all, but somehow I'm able to do it in Excel.

00:16:21   So I do what I need to do in Excel and then I save the Excel spreadsheet and then I import the Excel spreadsheet into numbers. And then the graph appears in numbers.

00:16:30   And I still can't even reverse engineer it. I so don't know how to graph things. And I would say spreadsheets are also terrible at graphing things, especially time series data where spreadsheets have no idea what time is.

00:16:41   They just think, "Oh, this is the first column. This is the second column, the third column." No, but they're time values. They're separated by time. Can you graph them over time? And it's like, "Nope."

00:16:50   I know you probably can. Excel geniuses will tell you how. Anyway, I don't like them.

00:16:54   So what this graph shows is related to something we discussed. And Jonathan sent us this giant spreadsheet filled with all sorts of information about various different silicon things, CPUs, GPUs, so on and so forth.

00:17:08   We will put a link to... I put the spreadsheet, I copied most of the spreadsheet and cleaned it up a little bit and put it into Google Sheets.

00:17:15   And we will put a link to that in the show notes. It's publicly viewable so you can look at it. It's way more data than we're going to go into here.

00:17:20   But the one graph I wanted to pull out was related to our previous discussion is how much bigger can Apple make the system on a chip?

00:17:27   And we were trying to look up die sizes and we didn't know them off the top of our head. And so this is why he was replying with all his info.

00:17:32   So here this graph shows... And someone should paste this into the chat room for the poor chat people.

00:17:37   It shows the die sizes of lots of different CPUs and GPUs and stuff. Way at the left end of the spectrum here is the A14 and A13.

00:17:47   And the A14 is actually smaller than the A13 through the magic of process shrink, even though it's got more stuff in it.

00:17:53   And the A14 is 88 millimeters square. Way on the right side is like this weird data center machine learning GPU that's just massive.

00:18:05   And it's 826 millimeters square. But right next to that, the second to biggest bar in the thing at 700, that's the 28 core Xeon in the Mac Pro.

00:18:17   So the A14 is 88 millimeters square and the 28 core Xeon is 700. So there's a lot of room for these chips to get bigger.

00:18:27   The M1 is around 120. This is an estimate. So from 120 to 700, that's currently the range from Apple's smallest Mac chip to its biggest one.

00:18:37   Actually, there may be an Intel one that's smaller, but I doubt it. So how much bigger can the M1 get?

00:18:43   Especially with regards to how much cooling capacity can it absorb? Well, if you've got a 28 core Xeon in your Mac Pro and it's able to cool it with that fan,

00:18:53   boy, you can make the M1 a lot bigger and a lot hotter before you run into any cooling limits.

00:18:59   It is darn near impossible to get an image out of Google Docs. My goodness.

00:19:03   Copy the link to the, copy the link to the, what do you call it? Because that has the image in it.

00:19:09   Just screenshot it and then paste it into a spreadsheet and then save that into a Word document.

00:19:13   Yeah, I put the Google Docs thing in. The Google Docs has, it has the Google Docs version of the chart, which I also probably wouldn't know how to do.

00:19:19   But because I imported it again, it figured, I just, I don't like graphing. Actually, I do like graphing. I just don't like these programs.

00:19:27   The way they think about graphing is not the way I think about it. So we clash.

00:19:31   Andrew Bunner writes, "I work in the computer vision space and one thing that we think about is how we can make giant ConvNet training jobs faster.

00:19:39   What on earth is ConvNet?" I'm assuming that stands for convolution, but who knows?

00:19:44   Okay, good talk. Here is Sarah Brass and they claim to have made a single die ML accelerator that is about eight and a half inches square.

00:19:52   I have no idea what the yields are or if it is even good, but the chips is an example of how big these things can be.

00:19:58   It's 56 times the size of the largest GPU. And so the largest GPU is 815 square millimeters with 2.1 billion transistors.

00:20:10   21.1 billion.

00:20:11   Thank you. Sorry about that. 21.1 billion. That's totally what I said.

00:20:15   Sarah Brass is 46,225 square millimeters with 1.2 trillion transistors and 400,000 cores. My word.

00:20:24   You have to look at it. It's basically like, you know the silicon wafers you see where it's like a big circular thing and then on it they print a bunch of different chips?

00:20:31   This is like just use the whole wafer for one, quote unquote, one chip.

00:20:35   And I mean, obviously it's not one chip. Obviously it is many of the same things over and over and over again, but it's hilarious.

00:20:40   Can you imagine if this was your target and every single one of them had to work? The yields must not be good.

00:20:46   But if you want to know how big can you make a silicon thing, you can make as big as the whole wafer if I guess money is no object.

00:20:53   And obviously if the cores are simple enough or whatever. But yeah, 46,000 millimeters square.

00:20:58   So again, M1 is 120. Big Xeon is 700. This one is 46,000.

00:21:04   That is utterly bananas.

00:21:06   All right. This is a wall of text, John. Do you really want me to read all this?

00:21:10   I mean, I can take a shot at it by paraphrasing because I think I understand the major points. This is a lot of tests.

00:21:16   This is from Wade Dragasses and he's got a lot of info about memory bus.

00:21:21   And we've talked about memory in the M1 and the unified memory architecture and bandwidth and so on and so forth.

00:21:26   So his first bit is he says, I've seen it repeatedly stated the M1 has 60 gigabits per second of memory bandwidth and that it's three times that of any MacBook Pro.

00:21:34   The M1 does in fact have 60 gigabits of memory bandwidth, but the 16 inch MacBook Pro has 40 gigabytes of bandwidth.

00:21:41   So it's about 1.5x, not 3x. And he gives all the math on this thing and it looks like it makes sense to me.

00:21:46   So I think people doing a 3x were maybe just not keeping in mind that a double data rate memory sends on the leading and trailing edge.

00:21:53   So you have to double it. Right. So that's why it's not 6x.

00:21:56   So it's not to diminish the M1 again, they just in the cheapest MacBook Air you can buy it has 60 gigabytes per second memory bandwidth, which is faster than the 40 in the most expensive Intel MacBook you can buy.

00:22:09   So what's novel here is the frequency being unpackaged and not having all the single-energy change latency.

00:22:14   Apple's been able to increase the DRM bus frequency by 220 percent versus 16 inch MacBook Pro.

00:22:20   While this increased bandwidth is mostly the most touted consequence of this also reduces memory latency and that's more likely the main contributor to performance.

00:22:27   Right. So talking about more of the benefits of the of the RAM being unpackaged, you can clock faster, it's closer, quote unquote, closer by and you have less latency and so on and so forth.

00:22:36   So I know some more speculation about what do you do in the Mac Pro and the iMac Pro when you need more memory.

00:22:45   So first on the topic of like, wow, the unpackaged memory has such amazing bandwidth or whatever high end CPUs today typically have at least eight channels of DDR4 offering 400 gigabytes per second of bandwidth.

00:22:57   So the big Mac Pro and other big computers have way more memory bandwidth than the unpackaged memory.

00:23:02   Unpackaged doesn't make it magical. Like if you throw money and buses at it and huge banks of RAM, you can get large amounts of aggregate bandwidth from doing that.

00:23:10   Right. And soon DDR5 will be out, which will basically double the bandwidth. Again, I'll put a link in the show notes to the Wikipedia page for a memory interface bandwidth numbers.

00:23:22   It says, I doubt Apple can push the frequency much higher than likely limited by signal integrity, if not thermal.

00:23:27   So to get into the ballpark, they'll need to increase the number of channels and therefore chips about eight fold.

00:23:32   But by using regular DDR instead of LPDDR, which is what they use in the unpackaged RAM, that amounts just eight channels and chips, very possible from all angles.

00:23:40   So we're saying it's basically if you put right now, we have those two little unpackaged chips in the M1.

00:23:47   Say you make room for eight of them. Say you put like, I don't know, put them all around the edges of whatever.

00:23:53   That's not a ridiculous or stack them or something.

00:23:56   Yeah. And they're talking about 3D stacking a little bit. But yeah, it's not ridiculous that they could do that.

00:24:02   That still leaves the capacity angle. It's likely impractical to fit terabytes of RAM onto the CPU package for the foreseeable future.

00:24:10   But I think there's a presumption being made that Apple cares about large memory systems. This is the point that Marco brought up in a couple of past shows.

00:24:15   Apple trumpeted the 1.5 terabyte support of the current Mac Pro because they could. It came for free from Intel. Well, not free, but you know what he means.

00:24:23   How much do you think that capability actually means to their customers? So Wade says is what I actually suspect is that they'll simply go with on package only, placing the capacity limit on something on the order of 128 gigabytes.

00:24:35   And I think the performance will be astounding thanks to the same kind of three times higher clock frequencies, et cetera, not to mention all their performance advances from the rest of the system.

00:24:44   So that's like this is where we're talking about. Is it feasible? How much could you get on package if money is almost no object? 128 gigabytes, probably feasible, but it would be really fast, 128 gigabytes.

00:24:56   And here is the here's Wade is obviously well steeped in the ways of Apple. Here is his justification.

00:25:02   I believe that Apple believes that they can get away with this by focusing on the end results, other domination over Intel and any and every performance benchmark, which they can trivially achieve from anywhere they are.

00:25:11   I mean, they're already achieving with the freaking MacBook Air, so no problem there.

00:25:15   If pushed, they'll use their typical line of we just couldn't find a way to achieve the performance we wanted by doing it the old fashioned way.

00:25:22   I find this plausible and in keeping with Apple's attitudes.

00:25:28   I don't think it's a slam dunk just because the Mac pros whole role is to have tremendous capacity and limiting that just cuts you out of the as Marco said in the last show,

00:25:38   just totally cuts you out of certain markets where people just need more RAM.

00:25:42   But those rumors of a half size Mac pro or like an iMac pro that maxes out at 128 gigs of RAM that has these characteristics, that would be fantastic.

00:25:52   And that is, you know, an extremely Apple like and a very likely machine that they could make.

00:26:01   So that was the end of Wade's comment. I have a few of my own to add for we've got so much feedback about here's how I think we can make bigger and better and faster max based on ARM chips.

00:26:12   Lots and lots of people. This is the most popular thing that people suggested is how about you just put two M1 chips in there.

00:26:18   Remember when Max used to have multiple CPUs? I mean, how about power Mac G5 had two of them? My Mac pro had two of them.

00:26:25   Why don't you just stick a second one in there? I think this is unlikely for a couple of reasons.

00:26:32   I mean, it is it is related to my next line item, which says low effort to Mac pro.

00:26:38   Some people think this is low effort. Why don't worry about all this of making the M1 bigger.

00:26:42   Just make like a little bit bigger M1, like an M1 X and use that in like your iMac or whatever.

00:26:47   And then on the Mac pro, just put two of them or four of them in or something.

00:26:50   Symmetric multiprocessing where you have multiple CPUs that each have their own caches, but share a central pool of RAM, like the setup that had been used on past Macs.

00:26:59   It's kind of a pain because as you can imagine, we talked about the cache hierarchy last time.

00:27:04   All these CPUs have to sort of keep on the same page, so to speak, pun pun, about what the deal is with memory.

00:27:11   So if they pull something from memory and it's in there like L3 or L2 or L1 cache or whatever, and someone else wants that piece of memory.

00:27:19   The CPUs need to communicate with each other and say, "Oh, you know, I'm in the middle of changing that."

00:27:25   Or, "You don't know that I changed that because it hasn't been written back yet, but I have a changed version of that in my cache."

00:27:29   Or if someone changes that piece of memory, they have to tell the other CPUs, "Hey, if you have this in your cache, invalidate it because I just changed it."

00:27:35   And that process, there's overhead to that process. It's not particularly efficient.

00:27:41   It's the same overhead when you have multiple cores and you keep jamming into the same chip, not the same overhead, but it's a similar type of thing.

00:27:47   But the more integrated the thing is, the easier it is for you to make a very efficient way to do that.

00:27:52   That's why within a single CPU, it is easier to deal with cache invalidation than it is with two entirely different chips on separate parts of the board that have to do that same task, but they're farther apart, have more latency, and they can't optimize the way they do this.

00:28:05   Especially if it's like, you know, two-way or four-way, now you have to handle even more cases. It's kind of a pain.

00:28:12   So I don't think Apple do that because I think the performance is worse, and I think the complexity is pretty annoying. It's not actually that low effort.

00:28:20   And speaking of low effort, I was thinking of this last show. This is one of the first notes I put in for next week's show.

00:28:25   What does a low-effort Mac Pro look like? I don't mean this to be insulting, but like, what if you needed to make a Mac Pro to serve the same needs as the current Mac Pro, but you just didn't want to spend some ridiculous amount of money to do the cool things that Wade was saying,

00:28:38   where you make this giant chip that has, you know, 128 gigs of unpackaged RAM and this amazing performance, and it's just this amazing bespoke beast that just has inside of it little bits from like the A14, but in general is this huge beast.

00:28:50   What if you didn't want to do that because it's just too much money? How do you make a Mac Pro with sort of the parts on hand? Let's assume you have something that's like an M1X that is beefy enough to be in an iMac or whatever.

00:29:02   Or you're willing to do that level of work. The low-effort Mac Pro is rip out the unpackaged RAM and use an AMD GPU and ship it, because you've got the M1 quote-unquote system on its chip, the RAM is in a bunch of DDR DIMMs, and the GPU is from AMD, your trusted partner for GPUs, and you write drivers to the AMD GPU, and you use the RAM and those giant banks that have tremendous bandwidth because it's very expensive and there's too much money.

00:29:31   It's very expensive and there's 12 slots and there's huge DIMMs in them, and you're done. That's it. That's the low-effort Mac Pro. It would have better CPU performance than the current Mac Pro. It would have equal capacity.

00:29:41   And the new AMD GPUs are actually giving NVIDIA a run for their money lately, so you're basically done. That's a great Mac Pro.

00:29:50   One of the things that are bad about the current Mac Pro is the CPU is slow. The GPU is actually as good as whatever you can buy and stick in there. It's as good as you can make it. Apple does not have anything that competes with the top-end NVIDIA AMD GPUs.

00:30:05   And the integrated GPU, as amazing as it is for an integrated GPU, does not compete with the big external ones.

00:30:12   The low-effort Mac Pro is still an option. We keep talking about all these fantastical things they can do, but they can also make a low-effort Mac Pro. I don't think it's a bad machine.

00:30:22   I think it's a machine that has a cooler, faster CPU, but has all of the same capacity. And thanks to the wonder of interchangeable parts and slots, the boring old "Hey, I've just got a bunch of PCI slots," suddenly you can put a 6800 XT in there, which is an amazing new GPU from NVIDIA.

00:30:38   And Apple likes AMD, and they just have drivers for it, and you're done.

00:30:44   I know the whole thing we've been talking about is Apple's never going to use the... They're only going to have Apple GPUs from now on. They're only going to write their own drivers.

00:30:50   Big Sur doesn't even ship an ARM version with any drivers for any GPUs other than the Apple integrated one. Like, the writing is on the wall that you're never going to see a third-party GPU in an Apple Silicon-based Mac.

00:31:03   We don't know the answers to those questions yet. All I'm saying is the low-effort Mac Pro is right out there as an option, it is entirely technically feasible, and I think it's actually still a good product, even if it's a little more boring.

00:31:15   I think the lowest-effort Mac Pro would be just keep it on Intel.

00:31:21   Boo!

00:31:22   And I know this is not the solution anybody wants, and I'm pretty sure Apple's not going to do this.

00:31:28   No, no. They said they're going to fully transition. The transition will take two years. They didn't have an asterisk that said "Accept the Mac Pro," which will never transition.

00:31:34   Right, exactly. So, if they hadn't said that, I would possibly think this might be even further out. But yeah, they did say that. But truly, the lowest effort would be just keep shipping Xeon workstations at the high end.

00:31:48   But again, that's not their style, and that's not what they're going to do. I do also think that they're going to keep having slots, and they're going to keep having interchangeable modular architecture.

00:31:58   And so, the way you do that is not by making some giant monster CPU that also has a bunch of GPUs on it.

00:32:07   The way you do that is by retaining some amount of modular expansion slots, whether it's RAM or cards or both.

00:32:14   Most of the feedback here was actually very interesting to me, because most of the feedback was way above my head, like way above my knowledge and expertise from people who really know what they're doing in this area.

00:32:26   And it was really, really nice to hear quite how large these other processors in the industry are that we use, or that people can use.

00:32:36   And so we know now that Apple can make a really giant M series chip if they want to that has tons of CPU cores, tons of GPU cores, as long as you can pay for it.

00:32:50   I mean, like those, the 28-core Xeon W that's in the Mac Pro is like a $3,000 or $4,000 CPU.

00:32:58   And granted, those are like Intel's prices, there's some profit margin there that Apple would be able to absorb into their total product price, but still, you're talking thousands of dollars in manufacturing costs just for that CPU.

00:33:11   But while most M computers, like most of the M-based Macs, I think are going to have no expansion possibilities whatsoever internally, like you're going to see similar things to the M1, soldered on RAM that's on or near the package, only GPUs that are built into the package, that are built into the, actually on the chip.

00:33:33   I'm not expecting to see any kind of, you know, other thing than that. Now that I've heard from all these listeners, and now that I see how big chips can get and still be part of like kind of mainstream products, I now no longer think they're going to kick the GPU off the CPU.

00:33:49   I think GPU stays on, and the only question is how you physically interact with memory in a short, fast way.

00:33:57   Even on the iMac, you think that?

00:33:59   I think so too. I think even on the iMac, I think we're going to have one giant chip.

00:34:02   I think seeing the benchmarks of the M1 in various scenarios, because people are just really benchmarking the heck out of it these days, the M1 wins everywhere until a real GPU comes walking along.

00:34:15   You're like, well, who cares about that? I don't care about gaming. There's a bunch of video apps that do a lot on the GPU, and you'll take a slower, crappier Intel Mac and put a decent GPU in it, and it suddenly becomes twice as fast at like 8K video rendering or whatever, right?

00:34:31   So, you know, it doesn't mean that Apple won't make that GPU, but they don't have anything that competes with that so far, and I do wonder if they have the will to do that, and especially given AMD's new GPUs, because when it was just Nvidia, it seems like Apple's just totally on the outs with them.

00:34:47   They're not going to do anything, but the new AMD GPUs and the new architecture, it's the same architecture that the new consoles use as well, but bigger and beefier, they're really good, and so in a top-end iMac, if someone buys that, they're not going to want it to be slower than their current iMac, and a lot of the tasks are GPU-bound, so I'm pretty well convinced that there has to be a discrete GPU in the top-end iMacs.

00:35:11   Otherwise, the ARM-based ones will be slower than the current iMacs for these kinds of tasks. Now, again, Apple could make that GPU, but we don't even have a hint of them doing that, so I'm leaning pretty heavily towards a discrete in the top-end iMac.

00:35:25   Yeah, I mean, maybe. I didn't realize there was that big of a performance gap, so yeah, that's very possible, but regardless, I think almost every other Mac has everything in or around the package, no separation of anything, and then only at the very high end, I think that's when you get separation, and it doesn't make sense.

00:35:46   Again, it doesn't make sense for the Mac Pro to exist if there aren't expansion slots, and Apple really, I think, very strongly communicated that that was the direction they intended to stay now that they rebooted the Mac Pro, they got it back on track, that Pro's like, "I think they're going to keep expansion slots,"

00:36:06   so I still think that's going to stay in that product, but I would be very surprised if any other Mac had any kind of expansion or nearly any kind of separation between the components at this point.

00:36:19   Final bit of follow-up on the making Macs bigger, faster, stronger. Something to keep in mind with regard to die size, cooling, and everything else is that TSMC, who manufactures Apple's system-on-the-shifts, TSMC is not Intel. TSMC continues to make process advances.

00:36:38   The latest story was about the TSMC achieves major breakthrough in two-nanometer manufacturing process, so to recap, the A14 and the M1 are at five nanometer.

00:36:49   Two nanometer would make for even smaller and more transistors and even less area. It would be pretty amazing.

00:36:56   This headline confused me because someone sent this link and was like, "They achieved major breakthrough in two-nanometer manufacturing process, risk production in 2023."

00:37:09   You know in headlines when you leave out articles or whatever, you leave out certain words for compactness, for the tradition of trying to fit things on the printed page?

00:37:16   Are they saying they risk production with parentheses around it as a compound noun in 2023?

00:37:27   What is risk production? I looked it up and found a couple of definitions. I'll put a link in the show notes.

00:37:32   I think one of the plain English ones from Tunafish on Hackers News was, "Risk production means that the foundry says, 'Okay, we think everything is fine now, but we make no guarantees that it will work.'

00:37:42   Customers then have the option of purchasing super-expenser wafer starts." You have to look up what "starts" means because it's in a whole other lingo thing.

00:37:48   "Super-expensive wafer starts that might or might not work. You might luck out and have lots of great next-gen chips months before your competitors, or alternatively, you might get five working chips out of each $10,000 wafer."

00:37:59   So that's what risk production means. So 2023 is the time horizon when conceivably, if you wanted to be on the bleeding edge, this would not be Apple, although for really low volumes, who knows?

00:38:11   Conceivably, at very low volumes, at high risk, you could get 2nm parts out of TSMC, which means that the 2024 Mac Pro, all the numbers we gave you about die sizes and everything like that, that's why in the graph where we pointed out that the A14 is actually smaller than the A13 despite having more stuff in it.

00:38:33   Again, 5nm shrinking down to 2nm doesn't mean they're going to be half as big for the same amount of transistors, but it means they're going to be smaller.

00:38:41   So, you know, we're back on the Moore's Law curve, because Moore's Law will end, we've discussed this before, but whatever curve we were on with Intel, we're not on that one anymore.

00:38:53   So expect every couple of years to have, maybe for the next couple of rounds, to have a process shrink and to have that process shrink ripple across Apple's products and really sort of give us continuing benefits.

00:39:08   That was another one of the questions of a lot of the feedback we got was like, "Can we expect this level of advancement with the M2, M3, M4?"

00:39:16   The easy answer is no, because you went from Intel that was stagnant to ARM, which is not, but you also shouldn't expect the M2 to be like 3% faster than the M1.

00:39:27   We're on a better curve than we had been, but there's not going to be another leap where the M2 is 7 times faster or whatever, right?

00:39:34   So things are looking up, but keep your expectations in check, but don't forget, shrinks will happen and shrinks are good.

00:39:41   Do you think a wafer start is kind of like, you know how I think, don't the silicon wafers start out as like a big log and they get sliced?

00:39:49   Do you think a wafer start is like the butt of the bread, but for the silicon log?

00:39:54   No, I think it's more like a sourdough bread starter.

00:39:56   Oh, okay, where you have to grow it for a while, feed it additional silicon every day?

00:40:00   It's neither one of those things. I didn't put the link in the show notes for wafer starts or starts in silicon manufacturing, but you can look that up as well.

00:40:08   Risk production was a new one for me, that's why I put it in the link.

00:40:11   One final bit about ARM Macs, this is not about making them faster, but just to note the Alexander graph, this story has already gone around a lot today, but from earlier in the week,

00:40:21   has successfully virtualized ARM windows on Apple silicon using QEMU plus a bunch of patches.

00:40:27   So, as you know, Microsoft does not officially support any of Apple's ARM Macs with Windows, despite the fact that Microsoft itself has a version of Windows that runs on ARM and they have their own hardware that runs on ARM.

00:40:41   But their hardware sucks because it doesn't use Apple silicon.

00:40:44   And so if you run ARM windows on an M1 Mac in this QEMU virtualization environment, it's about twice as fast as the fastest actual native Windows ARM hardware you can buy because Apple silicon is better.

00:40:59   So it's still, you know, as we said back in the ARM Mac introduction, the ball is in Microsoft's court.

00:41:07   Technically speaking, there is nothing stopping Windows from running natively on ARM Macs except for Microsoft, which has thus far not decided to support them.

00:41:16   I feel like Apple wants Microsoft to support Windows on Macs just because for the same reason Apple put that time and energy into boot camp, because it's something some of their customers find valuable and it makes Macs more valuable and more versatile.

00:41:28   So I hope those two crazy kids work this out because I really do want Windows on ARM Macs.

00:41:35   We are sponsored this week by Bombas.

00:41:38   Maybe you haven't always thought of socks as the perfect gift or the perfect way to give back, but actually Bombas socks were made to give, literally.

00:41:46   When you give a pair of super comfortable Bombas socks, you aren't only giving somebody a gift they'll love, you're also donating a specially designed pair to someone in need.

00:41:55   Because for every pair of socks Bombas sells, they donate a pair to someone experiencing homelessness across the US.

00:42:01   Since socks are the number one most requested clothing item in homeless shelters, the generosity of giving Bombas will make a meaningful impact this holiday season.

00:42:10   Their customers have already allowed them to donate over 40 million pairs of socks and counting through their nationwide network of over 3,000 giving partners.

00:42:18   And Bombas are really good socks. I'm wearing a pair right now because it's cold here today and I love their Merino wool line.

00:42:27   They have all sorts of great socks. Athletic, dress socks, standards cotton and everything. And they're all great, but for cold days in the winter I love the Merino wool.

00:42:36   It's a super warm, comfortable fiber and I usually itch with low quality wool. These don't have that problem at all. No itching, they're just super high quality.

00:42:45   I love Bombas socks. And they also design out that annoying toe seam, you know like that ridge in the front.

00:42:52   They moved the design around so they don't have that and they don't twist around your foot as you're walking to put the ankle in the wrong spot.

00:42:59   You know, you never slip, they have good grip, they have this cool little midsole support system. They're just great socks.

00:43:05   And they're 100% backed for life. If you aren't happy with them, just reach out to their support team and they will issue an exchange or a refund.

00:43:11   So from comfort to kindness and everything in between, Bombas aren't just giveable, they were made to give.

00:43:18   Go to Bombas.com/ATP today and get 20% off your first order. That's Bombas.com/ATP.

00:43:27   Bombas.com/ATP. Thank you to Bombas for keeping my feet warm, giving over 40 million pairs of socks to people and for sponsoring our show.

00:43:40   Marco, we haven't had a chance to talk about your new toy and oh, the time, the time. We do have to do Ask ATP. That stinks.

00:43:48   Well, there was. I did push one follow-up item about multi-chip modules down to next week, but I think you can't avoid it any longer, Casey.

00:43:55   The public needs to know about the iPhone mini.

00:43:58   Well, if you'll notice, what I named this topic was Marco's Mini Reviews, plural.

00:44:04   Oh.

00:44:05   You got the Mac Mini too, right? I get it.

00:44:06   No, not yet. We'll see.

00:44:08   Oh, you don't?

00:44:09   Apparently, the build-to-order Mac Minis are currently saying that they're going to ship in January, in late January at that.

00:44:14   So, yeah, maybe that plan is not going to happen.

00:44:18   But, no, my Mini Reviews are for my iPhone 12 Mini and my HomePod Minis.

00:44:25   Oh.

00:44:26   That's right. So many, many things. Or as Merlin would say, many, many things.

00:44:30   Exactly. Oh, poor Merlin. I'm from Ohio too. I get it. Just not that part of Ohio.

00:44:35   Anyway, the HomePod Mini is where I'll start. Everybody seemed a lot more excited about this product than I think is ultimately deserved.

00:44:47   And it's not to say that it's a bad product. It's a fine product.

00:44:51   There are a couple of places that I have enjoyed HomePods.

00:44:55   Let's say the smallest rooms in my house.

00:44:59   For the past hour long, the HomePod's been out about a year and a half or two and a half years, whatever it's been.

00:45:06   I've enjoyed the HomePod for that purpose, but it's a bit big and it's a bit expensive to waste in the smallest rooms of your house.

00:45:12   And so the HomePod Mini is actually really good for that.

00:45:16   I have swapped it in for that role and it is fantastic for that.

00:45:21   And it's also good in contexts where what you're mostly going for is the voice assistant and not necessarily music listening or where the quality of the music being played isn't that important.

00:45:35   But I don't think anybody with that list of qualifications is buying HomePods of any size.

00:45:41   The whole reason you buy HomePods is that they sound pretty good and if you really wanted the voice assistant part as the primary role of it, you're probably going to want an Alexa device instead because they're generally better at that.

00:45:53   Anyway, I had two HomePod Minis as a stereo pair, two HomePod Maxis, I guess, as a stereo pair, and also a new Amazon Echo.

00:46:04   And then I also tested out all of them as, I only had one Echo, I also tested them out as just mono single speakers.

00:46:12   Just by unplugging one of each thing and configuring the stereo pair and testing it just as mono.

00:46:19   The HomePod Mini is this tiny little ball about the size of a grapefruit. It's a hundred bucks.

00:46:26   The current, which is actually a brand new generation of hundred dollar Amazon Echo, the main middle of the road Echo, is also a hundred bucks, also a ball, but much larger.

00:46:38   More like it goes from a grapefruit up to almost a little bit bigger than a softball, I think.

00:46:43   Grapefruits are bigger than softballs, right?

00:46:46   It isn't a big grapefruit. It's one of the smaller ones. Maybe a large orange.

00:46:52   Is the Amazon one like a crystal ball, like a fortune teller?

00:46:55   Actually, it's almost like the original Magic 8 Ball.

00:46:59   Grapefruit and softball were wrong. It's more like large orange and Magic 8 Ball.

00:47:06   Is it bigger than a bread box?

00:47:08   Anyway.

00:47:09   What the hell is a bread box?

00:47:11   Whereas the HomePod, the full-size HomePod, is about two-thirds the size of a trash can Mac Pro.

00:47:18   If only there was some common unit of measurement we could use to express the size of things in a way the listeners couldn't understand. Oh well.

00:47:25   Well, you say that, but then we would use inches and everyone would complain and moan about how backwards Americans are.

00:47:30   Yeah, it's a millihome pod. A kilohome pod.

00:47:33   We'll do it in Planck lengths. Plank, Planck, I don't even know how to pronounce it.

00:47:37   Anyway, what you might expect is that the sound quality goes up proportionally with price, and it doesn't.

00:47:46   What you might expect instead is the sound quality maybe goes up proportionally with size.

00:47:52   Because really when you have HomePod Mini, Amazon Echo, Ball, and HomePod Maxi all in a row, it is kind of like small, medium, large.

00:48:00   It's a nice size ramp.

00:48:02   And that is pretty much true. That is effectively how the sound ramps up.

00:48:08   Because speakers, same thing with microphones by the way, you really can't defeat physics after a while.

00:48:17   You can play some tricks, you can be really clever, but if you're trying to have a point source emitting sound that is not extremely close to your ears,

00:48:28   you're going to have a hard time making that sound good if the point source is really small or really far away from you or a very small source in a very big room

00:48:39   or some other kind of large difference in the physics between the ratios or the proportions or the distances that you're trying to go.

00:48:47   Or it doesn't move a lot of air, right? Is that another factor?

00:48:50   Right, yeah. Sheer displacement here. Because you're dealing with a physical thing.

00:48:54   Sound waves are physical compression waves. At some point, physics is going to win.

00:49:00   I don't have a current gen Echo Dot, but I bet if I had a Dot, I bet it would fit in right below the Mini in size and sound quality,

00:49:11   but the Dot costs less than half or about half of the Mini's price, and I bet it isn't half as much worse.

00:49:17   Similarly, the Amazon Echo costs exactly the same as the HomePod Mini. Actually, when it's as it frequently is on sale, it costs significantly less.

00:49:27   And the new, what did I say, Magic 8 Ball sized Amazon Echo sounds significantly better than the HomePod Mini.

00:49:37   So it's the same price, it sounds way better, and the voice assistant is way faster and way more reliable and way better.

00:49:45   So ultimately, if you're looking to spend $100 on a voice assistant speaker thing, the Amazon Echo is a way better choice overall than the HomePod Mini.

00:49:57   That being said, the HomePod Mini is a HomePod, and therefore it uses Siri instead. To many people, that's a feature on a bug.

00:50:04   To me, it's kind of a wash. They're both idiots in different ways.

00:50:08   It's like the three of us.

00:50:10   So it's not an amazing home run product. And if you compare the sound from the Mini to the full-size HomePod, it's clear as day.

00:50:23   Like, oh yeah, the full-size HomePod sounds way better, especially where the Mini especially falls down. It's a massive difference in bass.

00:50:30   The Mini basically has no bass whatsoever, and the regular HomePod has surprisingly good bass for its size.

00:50:36   Because the regular HomePod is still, in the world of speakers, is still small.

00:50:40   In the world of smart speakers, it's maybe medium to large, but in the world of speakers as a whole, it's quite small.

00:50:46   And it competes very well with stereo speakers that are much larger than it, especially when you have two of them.

00:50:52   So I think ultimately, the regular HomePod sounds like it's twice as good, and it's about twice the size.

00:51:01   Unfortunately, it's nearly four times the price officially. And so I think if you can actually have the HomePod, the full-size HomePod, at its frequent sale price of $200-ish, that's a fantastic buy.

00:51:16   And that is totally rational and sensible as twice as much as the HomePod Mini, because it's about twice as good.

00:51:23   As for the HomePod Mini itself, like where it's left, it's kind of, you know, the size kind of suggests its quality pretty well compared to the other ones.

00:51:32   If you have a roll in your house that you want to fill with a very small smart speaker that is not that expensive and is not that good sounding, but where that might not matter or you might not care, it's a good buy.

00:51:48   I'll give it that. But that's a lot of ifs. And if you want something to like, you know, play music in your kitchen or something, the Echo is going to be better for that or the full-size HomePod.

00:51:59   As with the other HomePod, if you can afford two of them, it is great. It is way better than just one.

00:52:07   However, even in a stereo pair, the HomePod Mini is not a real winner with sound. It doesn't make up for its lack of bass or the rest of it being kind of okay.

00:52:19   Whereas like a stereo pair of the full-size HomePods really does sound quite good in almost any room.

00:52:26   I would strongly suggest if you have, you know, a roll in maybe the smaller rooms in your house where you're mostly playing podcast and you don't care about sound quality, yeah, go for it. It's nice. It's inexpensive.

00:52:38   If Apple would have made the HomePod Mini the size of the current Amazon Echo instead, it wouldn't have been as small.

00:52:48   But people don't really need these things to be super small. So did they make it super small because Apple, even though nobody was really asking for it to be quite this small if it's going to sound quite this mediocre?

00:52:59   Or did they make it this small because at $100, their profit margins wouldn't allow them to give it like two drivers?

00:53:09   Because one of the reasons it sounds so bad is that it doesn't have a separate woofer and tweeter. It has only one driver and then it has these "passive radiators" on the sides to kind of enhance bass response.

00:53:21   I don't know enough about speaker design to know what those do, but every other speaker in this price category has multiple speakers inside. There's usually a subwoofer and a tweeter or a woofer and a tweeter at least.

00:53:34   And so if they prioritized size here, I think that was the wrong move. If they could have shipped something with better speaker drivers and maybe more of them at this price point, I think they should have.

00:53:47   But ultimately that could have just been a price decision as well. It's a fine product, but it's not competitive with the Echos at similar price points.

00:53:56   So a couple of things that I wanted to touch back on. First of all, you had made mention that you don't have an Echo Dot. We have year old Echo Dots, so it would not at all surprise me if these are now out of date.

00:54:08   But I can tell you with a fact, they sound like trash. Which is fine. If we're going to put on music on the Echo Dots, it's because we're trying to get multi-room going and we just want seamless music throughout.

00:54:23   And we only have these upstairs. So we just want seamless music throughout the upstairs. We are not listening, like what is it, the Maxell or whatever it was called that I've mentioned before, the advertisement with guys with his hair blown back.

00:54:35   We're not sitting there like listening and really paying attention. We just want some ambient stuff on. And for that it's fine. But they sound like garbage, unequivocally.

00:54:44   The other thing you said is that the HomePod Mini has no bass and that alarms me greatly because I am not that much of a "Ntntnts" kind of guy.

00:54:55   But my understanding of your music preferences is that they are not particularly bassy, generally speaking.

00:55:02   And so if you're saying with your comparatively tinny music preferences that they're not very bassy, someone who likes something that is less bish and more, I don't know, like bassy, that could be quite the deal breaker.

00:55:17   Yeah, I mean again, if you care at all about music sound quality, this is not a product for you. Period. End of story, that's it. And generally in the world of smart speakers, they don't sound great.

00:55:33   Because again, part of the physics mentioned earlier is every one of these speakers and the whole industry of soundbars, which I also hate soundbars. What an amazing branding for such a stupid concept.

00:55:47   The soundbar problem, which is the same problem that almost every smart speaker has, is that they try to play tricks to defeat physics, to try to make a wider soundscape basically. To simulate having multiple speakers in the room when you're actually broadcasting sound just from one point.

00:56:08   And so you have to do things like try to bounce the sound off the walls or in the corners or whatever. And while that's a real thing, you can indeed bounce sound that way, in practice, it doesn't work very well in most rooms.

00:56:22   And it's not nearly as good as just having two different speakers that are just one at the left and one at the right. It's not nearly that good.

00:56:30   Similarly, all these speakers try to somehow sound okay while broadcasting from one single point in a room. Oftentimes that can't be the ideal space that you maybe want to put a speaker if you cared about sound quality.

00:56:44   So you're dealing with these compromised physics situations here. And so all smart speakers kind of have to be graded on a curve.

00:56:53   And so from that point of view, you could use the HomePod Mini and okay, commute and fun, fine. But if you're going for commute and fun, I honestly I would go for the Echo again.

00:57:03   Like unless you really love Siri over Alexa for a particular reason that's very important to you. So it could be privacy, it could be maybe some of Apple's integrations.

00:57:12   AirPlay 2 is a big one and I love AirPlay 2. And even though the latest release of iOS has made it incredibly buggy, thanks. Thanks HomePod Mini.

00:57:21   But if you don't have a significant reason to prefer HomePods specifically, it's not a great product. Because even by the grading on a curve that you have to do for smart speakers.

00:57:34   And even though the regular HomePod, the full-size HomePod almost eliminated that curve. Like you don't have to grade that on a curve. You can just grade it as a nice speaker and it does pretty well for its price point.

00:57:46   Especially at the discount price point. But the HomePod Mini just doesn't. Like it doesn't, it's not a great product. It's a cute small product that is fine.

00:57:56   But Apple I think cut a few too many corners on this one. And they should have gone a little bit higher. And even if they had to make it cost $129 instead of $99.

00:58:08   If it could have had significantly better drivers in there and maybe two separate ones, that would have gone a long way towards sound quality. And if it made it a little bit bigger, oh well, nobody cares.

00:58:18   I just want to add that while you were talking about this, my wife messaged me to say, quote, "Siri sucks." Now here's the thing. Is she listening to the show live and wanting to chime in about the fact that Siri is not as good a voice assistant as Amazon Echo? Or is she just living in my house where Siri exists and occasionally does things that annoy us?

00:58:41   There's no way to tell. Because very often people in our house will have complaints about Siri that are unrelated to listening to a live podcast. So yeah, Siri's still got some catching up to do.

00:58:51   Oh, and we've fired our Sonos stuff and it's so good. I'm so happy without the bug. Sonos makes some good products. The Sonos amp is awesome. I love the Sonos amp. It's a stereo speaker amp with AirPlay 2, with TV inputs and analog inputs. It's wonderful. I love my Sonos amp. That is what drives my TV speakers.

00:59:14   The reason I love the Sonos amp is that it doesn't have any smart integration. And where the Sonos products really fall down is the ones that integrate the Alexa support and try to make these combo voice controlled smart products. I don't know what it is about that integration, but it's buggy.

00:59:30   All the reviews, people were complaining about the Alexa integration being buggy. And I thought, oh, maybe that was just like, you know, beta 1.0 software issues. And maybe they've made it better now, like, you know, multiple years in. No, they haven't. It's still really buggy.

00:59:45   And the Sonos move, I feel the same way about like, we have one of those, I just want to sell it now. Like I want out of that entire ecosystem. Anything in that ecosystem that has a voice support, I want out of it. So I'm just going to be unloading those things probably.

00:59:58   But the Sonos amp is wonderful because it keeps it dumb. It lacks both the buggy Alexa support and any reason to ever use Sonos' apps for anything. So it's great. Like I just use it as my TV speaker and it's an AirPlay 2 destination. It's wonderful and that's it.

01:00:17   Well, that took a lot longer than I expected. So let's do some Ask ATP.

01:00:21   We have to do the iPhone mini. I think I actually might have less to say about the iPhone 12 mini.

01:00:26   Alright, go ahead.

01:00:28   Alright, so I got my kind of red iPhone 12 mini. The back of it is not even close to red, but that's okay. I never look at the back. The sides are red and they are delightful. And every time I see the aluminum sides of my phone, it just makes me smile. I love the way this phone looks.

01:00:49   In the hand, it feels fantastic. In the pocket, it feels even better. I have been using it without a case so far. I have the Apple case. I figure kind of, you know, maybe just like if I can ever travel again, I'm probably going to want a case. Or maybe I figured I might try it for like a week and see what I like better.

01:01:08   The screen size is great on the mini. You know, I don't feel like I've lost any screen real estate because I didn't. Like they did just shrink the 10, 10S and 11 Pro screen to fit this size now. So like screen real estate wise, it's the same. They just made the pixels smaller.

01:01:27   Now on the small pixel side between the 12 mini, which is shrunken the pixels and my new MacBook Air, which I'm running at its default resolution. Whereas compared to my 16 inch, I ran at its native resolution, which is larger pixels.

01:01:44   I'm looking at everything a little bit smaller now and I'm starting to feel, hmm, I'm 38 and a half now. I've never worn glasses, but I don't have many years left of not needing any vision correction and having good uniform vision.

01:02:05   I can kind of feel it slightly slipping away. And as I am using these two new devices that have more dense screens than what I was using before, I'm noticing that. I'm noticing like, hey, this is a little bit harder to read than it might have been a few years ago.

01:02:21   Even though what I've just done is go back to the density of all the phones before the 10. Like that's what this density is. It's the same density that the 6 and 7 had. But I am kind of thinking like this might be a one year thing because I might have issues with just looking at very small things within the next year or two.

01:02:45   You can crank up the zoom though, right? Didn't ask you during set up which size you prefer? Well, I can, but then I lose screen real estate and that I would be less willing to lose that possibly. But I mean, we'll see.

01:02:54   But you still keep the good hand feel and the good pocket feel, right? Yeah, maybe. I'm kind of hoping that maybe in future revisions they might reduce the weight of the middle Pro line a little bit because they're so dense and heavy.

01:03:09   And I feel like I got worse this year anyway. So otherwise, though, like it feels wonderful grip wise. I'm still getting used to not having a leather case because this is the first time I haven't had a leather case since the 5S.

01:03:23   I'm still getting used to like picking it up off with the edges and having I kind of still feel like I'm going to drop it all the time. So that's another reason why I might try the case. Another reason is that in my previous caseless days, the phone would rest flat because there was no camera bump.

01:03:40   Now there's a camera bump and so with no case the phone does not rest flat. Yeah, it's infuriating.

01:03:46   That's one of the main reasons I use a case these days. I demand that it both lay flat and not make me nervous about scratching it. But laying flat, yeah, that's why I really was upset by the Apple leather Pro Max case, which adds a bump on the bump.

01:04:00   So even when you get the case that doesn't lay flat, I was afraid that would be true of all the cases, but apparently not on the non-Max size of Apple cases and certainly not on the Sena case that I'm using. Laying flat is a major feature.

01:04:12   Yeah, exactly. And also just like having a little bit of that tackiness grip when you're on a surface. Like I think I'm ultimately probably going to go with a case, but I want to try cases a little while longer because it just feels so good.

01:04:25   And it looks really good. Again, like I love the red aluminum edge. I absolutely love it. I like it way better than any of the other colors.

01:04:34   Performance wise, battery wise, the battery is noticeably lower. Like as everybody said, yeah, I do get less battery life. It has not been a problem yet.

01:04:43   And ultimately if I'm going to like ever travel with this phone, which honestly it's probably not going to be a whole lot of that, but if I'm going to ever travel with this phone, I would probably want some kind of like MagSafe battery.

01:04:57   If Apple would make that, that would be awesome. I don't know. I don't think anyone else has made one yet, but that would be great. Like a little like thing you just stick on the back for charging in your pocket sometimes, but not all the time. That would be wonderful.

01:05:08   But that doesn't exist yet. So we'll see. I haven't actually tried MagSafe yet. I've just been using my regular Qi charger that I've been using for years like on my nightstand.

01:05:17   So I haven't tried that yet. Cause ultimately what I want for MagSafe is something that is stationary on the nightstand that like I don't want their weird little travel wallet they unveiled for way too much money yesterday.

01:05:29   Like I don't want that. I don't want just like the loose cable. I want like a dock that is basically a magnetic Qi charging stand like that or pad or whatever. That's what I want.

01:05:39   And I don't, I don't know that there's a lot of those on the market yet, but anyway.

01:05:43   That's been another issue with the mini with dock like things. If you have a dock that's made for quote unquote normal phones, the mini is so small that depending on your luck, it may be that the coils don't align as well because the mini is shorter. You know what I mean?

01:05:58   And so either it doesn't charge as fast because it's misaligned or in some of them it just won't charge at all because the thing thinks it's misaligned and won't charge it.

01:06:06   I don't think this is a big problem. Panzer posted about it when it first came out, but it is something to keep in mind because the mini is so different than the average size phone. If you just have a generic phone charger or even worse, a charger that is custom made to fit like a galaxy note or some other huge thing. Be careful and think about it before you buy it.

01:06:24   Yeah, so otherwise, like other than like, you know, the battery being a little bit low and you know, having having a dilemma on whether I want to use a case or not. Otherwise I'm loving this thing.

01:06:36   It does feel a little bit like going back in time in the sense that it is so much smaller and lighter and you do have less screen physical size.

01:06:46   And so for things like watching video and stuff, it's like slightly worse than the big phones. But ultimately, you know, we're talking about what a half an inch or something or like a one inch screen size difference.

01:06:59   That's a lot proportionately, but you're still watching video on a tiny screen. And so if you're trying to decide whether you want to watch video on a tiny screen or a tinier screen, like it's they're both massive compromises compared to like a TV or even a laptop.

01:07:14   So that almost doesn't matter at all. Camera wise, so far, I have not missed the 2x camera that much. I have taken pictures where I have zoomed in digitally, but usually not to the full 2x.

01:07:32   And the digital zoom has been good enough for my like, you know, 1.4 1.6 kind of rings. It's been good enough for that.

01:07:39   And there have been some pictures where, you know, we take pictures like around dark dinner table like we took our Thanksgiving dinner pictures around our dark table and Tiff had her max and I had my mini and her pictures did look better.

01:07:50   They did have a little bit more detail, but I had to like zoom in and like and really look for the difference.

01:07:56   And so from my point of view as as a much more casual iPhone photographer, like I don't I don't I'm not a pro iPhone photographer. I barely posted Instagram. I am, you know, I don't shoot raw. I don't use halide or any of the fancy apps that people who are really good at this stuff use.

01:08:12   I just shoot with the building camera and occasionally tweak the photos a little bit and occasionally post some of them to Instagram. But even then, it's pretty rare. And so from my point of view, the 1x camera is great and digital zoom when I need it so far has been okay.

01:08:28   If there were a 2x option on this size, no question, I would pay extra for that. I would definitely take that. But there isn't. And so ultimately, I do like the size a lot, even despite that.

01:08:40   Finally, the speaker seems to be like some of the reviews tried to say that the speakers were kind of comfortable between them. They're not the speaker on this one is a little bit lower volume and a little bit worse than the bigger ones, but it's not a big difference.

01:08:56   It's it's not a difference that I would consider fatal and I use the speaker a lot, mostly for podcasts because it sounds like an echo dot for music. So anyway, ultimately, I'm very happy with it. I'm a little scared that no one's buying it.

01:09:13   Because when I look at my stats in Overcast, there's not a lot of 12 mini compared to the other ones. So I'm a little worried about that this might be like a one off thing. And like they might not make another one next year or ever.

01:09:30   If that's the case, if this is not like a regularly updated product line, I'm not going to like hang on to this forever. I don't love it so much that I would do that. Like if there if next year, the lineup is similar, but without a mini option.

01:09:43   I'll just go back to the you know, the smallest pro like I have for the last few years. But if this is still an option next year, I'll probably take it because it just feels so good. It's so light. It's so small, but it's not too small.

01:10:01   It doesn't feel too cramped. I haven't had problems like with the keyboard accuracy or anything like that. Like, it doesn't feel too small. It's just right on that edge. But it's not. So ultimately, I'm very happy with it.

01:10:14   And and yeah, we'll see. We'll see how the rest of the year goes. We'll see like what I think about it. If I travel with it, I'm sure the battery life will be a hit there. We'll see you know how the lack of the 2x camera plays out over the year. But for at least this year, I'm very happy with it. I'm very happy with the almost red color that I picked. And yeah, I guess I'll keep you posted on my exit interview, at least at the end of the year to see what I think what I think then

01:10:39   I think I survived this. You're not making me feel that much FOMO, which we all know for me it's it's but but a small breeze that can that can get me right into the FOMO side of the world.

01:10:50   You haven't felt one yet. Yeah, well, just wait. I don't know so far. So far. I'm not I'm not feeling terrible, terrible amounts of FOMO. So I consider this a success.

01:11:01   We are sponsored this week by Squarespace. Start building your website today at Squarespace comm slash ATP, enter offer code ATP at checkout to get 10% off.

01:11:13   Make your next move with Squarespace. Now look, you're listening to podcasts, you listen to this podcast, you've probably heard of Squarespace ad before, and maybe just this one moment.

01:11:24   Maybe you don't need to make a website. But there will come a time when you need to make a website for somebody either for you for yourself, or for somebody else.

01:11:32   And there's so many ways to make websites these days. And many listeners of our show can code your own CMS, I can I've done it. But most of the time, that is not worth doing.

01:11:45   Most of the time you want to outsource the creation and the hosting and the editing and the support of websites to specialty platforms. And Squarespace is by far the best place that I've ever used to do this.

01:11:58   It is super easy, no matter what your skill level is, whether you're a programmer, or whether you're not, whether you have web dev skills or CSS skills, or whether you have none of that, you don't need any of those skills.

01:12:08   Because Squarespace is all visual. It's all what you see is what you get live previewing, everything's intuitive, drag and drop, super easy to use.

01:12:16   And you don't have to worry about software upgrades or security patches on the server or anything like that. Squarespace supports it and they keep it running.

01:12:24   So whether it's for you or somebody else that burdens off your plate. So start it, you can see for yourself what you can do there by starting a free trial site right now at squarespace.com/ATP.

01:12:35   Start a free trial. You can build the entire site without paying a dime or without giving them a credit card, just so you can see how it works.

01:12:42   When you decide to sign up, and I bet you will, go to back there squarespace.com/ATP and use offer code ATP to get 10% off your first purchase.

01:12:50   That's squarespace.com/ATP, code ATP for 10% off. Thank you so much to Squarespace for sponsoring our show. Make your next move with Squarespace.

01:13:00   All right, let's do some Ask ATP like for real this time. Matthew Taylor writes, "Is it possible to disable the fan on the new M1 MacBook Pro? Does Big Sur contain a setting for such or do you know of a third party app that can do so, thus forcibly restricting system-wide performance to a fan-off thermal envelope?

01:13:19   What annoys me about my current 2015 MacBook Pro is the fan noise. I'd love to be able to make fan usage an option I control."

01:13:25   I understand the question, but I have a few problems with this. First of all, anyone I have heard, read, spoken to who has one of these says that the fans almost never kick on, first of all.

01:13:36   Second of all, if you don't want fan noise, buy a MacBook Air. That'll fix the problem for you.

01:13:43   It's nearly the same computer. Oh, by the way, oh my god, just a very quick update on how I am liking my MacBook Air.

01:13:50   I love it so much. You know what I love so much about it? Not the color. The color I think was a mistake. I got the gold, probably a mistake.

01:14:00   I could have told you that.

01:14:02   Yeah, but it does do one thing though. It makes it look new and different.

01:14:06   Because while I know the gold has existed for a couple of years in that product and then for even more years before that on the 12 inch, although that was a different gold.

01:14:15   Anyway, I've never had it. So it's new to me. There's never been one in the house even. So that is new.

01:14:21   So to me, if you don't get a weird color on this product, you can't even tell what it is. You can't tell that it's this radical new cool thing.

01:14:31   And for many people that's a feature. But for me, I kind of felt like, you know, I kind of want this radically different new awesome thing to look different, to look new.

01:14:41   I'm so tired of Space Gray. Space Gray is, I'm way, way over it.

01:14:47   Silver is what I probably should have bought. But I've had silver laptops now for how many years? Like it's fine. It's a great color.

01:14:55   But it just looks, it made it look and feel older, I think. So I'm very happy that I got this weird gold, orange, pink, whatever it is color.

01:15:03   Because it looks different. Even though I actually don't like it in certain lighting. But in certain lighting I do.

01:15:10   You got to get some skins. No, those are, I've tried that. I've been down that road. It's not for me.

01:15:17   You put a paper-like thing on your iPad, why not put a skin on your MacBook Air? Although finding one that won't clash with the gold is a problem now. But you did that to yourself.

01:15:26   Yeah, right. Congratulations, you played yourself.

01:15:29   But one other thing, it doesn't have a touch bar. Oh my god, I'm so happy. I love, like, hey, you want to turn on the volume? Just tap. One tap, not two.

01:15:40   You don't have to wake it up first and then find where the button is and then drag it down. Nope. One tap on a button. Boop. Volume's down.

01:15:48   You want to increase the brightness? Oh, boop. One brightness.

01:15:51   Don't make people tell you about the tap drag gesture. Because they're going to. Don't make them.

01:15:55   I know about the tap drag. You still have to tap it to wake it up.

01:15:57   I know. I'm just reminding. I know that you know, but people don't know. I'm saving you.

01:16:01   I know. And also, by the way, don't write in to tell me what a bread box is. I know. My grandmother had one.

01:16:07   Yeah. Also, don't tell me about a better touch stool and how much better it makes the touch bar. I've tried that too. Still hate it. So, anyway.

01:16:14   Yeah, you know what? Can we explore that just briefly? Can we explore that briefly? Because when I got my MacBook Pro six months ago, which was the first touch bar to be in my own home.

01:16:23   First of all, I don't have particularly strong feelings about the touch bar. Like, I don't think it adds anything, but it doesn't actively piss me off like it does a lot of people I know, including you, Marco.

01:16:34   But I thought to myself, well, you know, I should use Better Touch Tool for doing cool stuff. And I think we spoke about on the show, you know, that one of the things I did was put like a little emoji representing whether or not the garage door was open on the touch bar.

01:16:46   Not because it needed to be there, but just because I thought it was a neat thing to try. And it worked. But I personally found Better Touch Tool to be extremely unreliable and very crashy and just did not work well with the touch bar at all.

01:17:00   Now, it very well could have been user error. And honestly, if it was, I don't care because I don't think I care enough to go digging and figuring out why this was the case.

01:17:08   I know other people have had very good experiences with Better Touch Tool with the touch bar, but for me, it just felt like a pile of hacks that just wasn't really for me.

01:17:18   And so those of you who are Better Touch Tool fans, I respect that, but I don't know how you do it.

01:17:25   Yeah, I mean, in all fairness, the touch bar is very buggy and crashy and unreliable, even stock. But yeah, I am so happy with this laptop. Not having a touch bar to me is like not having a fan. Like it's just it's wonderful.

01:17:39   Here's this thing that occasionally annoyed me and now it's just gone. Wonderful. And I was thinking too, like it's probably going to have a longer lifespan because there's no moving parts.

01:17:51   Like, you know, I'm having this problem with my iMac being filled with dust after three years. Laptops have that problem too. Many people frequently have problems with their laptops, like where the fans spin up like crazy way more than they did when they were new.

01:18:03   And some of that's, you know, because of weirdness and, you know, maybe thermal compound degrading, but some of that's also because stuff is really full of dust in there and it's hard to get it out.

01:18:11   And so when you eliminate the fan, you eliminate not only the noise and annoyance of that, but so many problems that could occur in the future as you use this machine for years and years that I think, you know, ultimately the main limiting factor of this is probably going to be the battery lifespan, not anything else about it unless, you know, it gets damaged.

01:18:30   Because there's just nothing about it to really go wrong. And that's yet another feature of the Air. And again, you look at the lineup and, you know, going back to this Ask ATP question that I really interrupted, sorry, Casey.

01:18:41   Like, I'm with you, like, if you're looking into the MacBook Pro, the current one, the 13 inch, you know, kind of mini MacBook Pro, and you're worried about the fan noise, just get the Air.

01:18:54   I mean, even though people have said that it's really hard to make the fan turn on, like, if that's going to be a main priority for you, just get the Air, because it's nearly the same computer.

01:19:04   Yeah. I don't think it's about the fan turning on. Like, I would, I don't recommend getting the Air if you want a MacBook Pro. If you want a MacBook Pro, get one.

01:19:12   Like, the reports basically say that even though the fan is spinning, it's spinning so slow that you can't even hear it. If you have a 2015 MacBook Pro and you're annoyed by fan noise, don't use that as a way to measure how you're going to feel about the MacBook Pro with the M1.

01:19:27   Just get the MacBook Pro with the M1 in it. All reports have said, you will, if someone just tells you it's fanless, you won't even know. Because even when people are running benchmarks for 10 minutes, it was like Gruber was saying he literally could not hear it.

01:19:38   He put his ear right up to the thing. Running a benchmark, he couldn't hear it. The fan is in there, and it is spinning, but it's spinning so slow that it's not going to be anything like your dust-filled 2015 MacBook Pro.

01:19:50   So if you want the M1-based MacBook Pro, just buy it. If you want to be sure you're not going to hear a fan, get the Air. But I don't want to steer people away, because all the reviews have been unanimous in saying, like, yeah, it's got a fan, but don't worry, you won't hear it. It's not like the fan on your current Intel MacBook Pro.

01:20:07   Also, to actually answer the question before we move on.

01:20:11   Oh, yeah, you can get software to turn off the fans. In fact, I forget which app I use, but I did it on my Mac Pro. I can stop all the fans on my Mac Pro, which is not advisable. You can watch the temperatures just go up, right? But yeah, you can do that, but don't. Please don't do that.

01:20:27   Now, I have a question. So the way those programs, like CoolBook, the way those used to work is they would interact somehow with the power management unit of the computer. I don't know what mechanism they could do that, but I would expect, first of all, I would expect all of that to be different on the M1-based Macs.

01:20:43   Possibly not to be available at all for software to modify, or if it is available, it will probably be a while before the apps are updated to support it.

01:20:51   Yeah, they update pretty quickly. Remember when we were going through this with my Mac Pro? It took a while for any app to support fan control, but eventually, because it's the Mac, you're not limited in the same way you are on iOS devices. So yes, it is an M1 and it's new and they'll have to come up with new stuff, but they will do it. There's nothing that's stopping them.

01:21:07   Unless they don't have access to it. Unless there is no API they can call to modify that on the M1 architecture, which is possible. Anyway, so I'm curious though, Jon, because the way those used to work is they were able to adjust the minimum speed so they could speed the fan up, but they couldn't slow it down beneath what the unit wanted to run it at.

01:21:30   If temperatures reached a certain level, the system would enforce its minimum speed no matter what you told these apps to do. So all these apps could do would be to increase the speed so that maybe you could, if you wanted to, the reason why it's called CoolBook, is you could run your laptop cooler by having the fan blow faster.

01:21:46   So you're saying that with your Mac Pro, it was able to actually lower the speed beneath what it would normally be?

01:21:52   So I didn't play the full game of chicken to see, hey, what happens if I just leave it like this? But you could put it into manual control. I think the app I was using was TG Pro. You could put it into manual control and it basically gave you sliders.

01:22:04   And you'd hear what the fan level is at now and you'd move the slider down and it would get quieter and then you'd watch your temps start going up. Right? So clearly it's whatever I'm doing, I'm making it be below what it would want to go because the temp would start climbing and climbing.

01:22:17   And then I'd just put it back into automatic because I didn't want to see what would happen. I didn't want to find out where the limits are on my very expensive computer. But I'm pretty sure that it really was bringing the fans down to below where they wanted to be.

01:22:31   It could be that the default mode is above the minimum. To your point, maybe there's a minimum and they don't run at the minimum. They run in the middle. But it did give me a slider and you could slide it all the way to the end. It did make it quieter and it did make the temperatures go up.

01:22:44   But please don't do this to your computer. Especially an M1 MacBook Pro. Just get it and just tell yourself there's no fan in it and see if you ever hear it.

01:22:51   As another quick aside from the Ask ATP that will never end, I recorded using prior sponsor channels. I recorded a couple of things, a couple of the holiday specials off ABC and I was running them through Final Cut Pro in order to take the commercials out.

01:23:07   And I rendered them, these are like 45 minute, maybe an hour, 10 minute total television programs and they were 720. The source was 60 frames per second which I did not keep and I cranked it back to 30 I believe.

01:23:23   But I was rendering them out of Final Cut Pro. There were no fancy transitions. There was no weirdness to them. I did render them as HEVC on my iMac Pro.

01:23:33   One of them took literally like six or seven hours. I have no idea why. Nothing else was going on in my computer that was abnormal.

01:23:41   And after that was done, I very nearly bought myself a new Magmini. This is preposterous.

01:23:50   Now, I know I'm a media pack rat. I know I have very esoteric and peculiar needs when it comes to media stuff.

01:23:57   It is unusual for me to be going into Final Cut Pro and doing these sorts of things. But I don't know why it took so long. People are already unloading in the chat.

01:24:06   What about hardware encoding? What about hardware encoding? Yes, it has hardware encoding. I know it does. I have no idea why Final Cut Pro, yes, it's not Final Cut Pro 10, it's just Final Cut Pro now.

01:24:15   I don't know why it took so darn long. But I started it around lunchtime and it didn't finish until around dinner time. I have no idea why. But, oh my goodness.

01:24:23   Was it using the hardware? Because the hardware is not infinitely capable and I'm not sure under what scenarios Final Cut Pro will choose to use the HEVC encoder that's in the Intel CPU versus not.

01:24:36   One of the big things is that if it has 10-bit video coloring, it can't use the T2.

01:24:41   And that may have been, in fact I think that might have been it now that you say that, but either way, I have encoded like H.264 stuff, I've re-encoded as HEVC in order to save storage space on the other end.

01:24:55   So if it's something that I'm, you know, it's not critical to me that it's in the original form factor in which I downloaded or the original codec I downloaded.

01:25:03   You mean ripped.

01:25:04   It depends on what we're talking about. I take your point, but it depends on what we're talking about. But nevertheless, I can do, I can usually on the iMac Pro get anywhere between 1x and like 3x with the CPU doing HEVC encodes.

01:25:19   And this was like 1/10x and maybe it was the 10-bit thing, I don't know.

01:25:24   Well, you'll know because if it's using the T2 acceleration, the CPU will be almost idle. Like you'll be able to see it in iStack menus. The CPU is doing almost nothing.

01:25:34   I don't know, whatever. Hearing all these people saying, "Oh, I can render in Final Cut Pro so fast." It was quite the sales pitch for me getting a new Mac, which hand to God, I do not want a new Mac.

01:25:47   I don't want a new Mac Pro right now. I probably will soon. I do not want a new iMac right now. Probably will soon.

01:25:52   All right, listeners, you heard it here. So which do you think happens first? Who caves, me or Casey?

01:25:56   Hold on.

01:25:58   Which of us has a different Mac or desktop first?

01:26:02   Well, so that's interesting. That's actually a very interesting question because I, sitting here now, I really don't have any need or desire, he says with little confidence, to change my iMac.

01:26:13   I really do think though, as soon as there's a 4.13 inch MacBook Pro, it's quite likely I'm going to pull the trigger real soon.

01:26:22   And you already did. I mean, you've already lost that fight in the laptop department. So I mean, I already won.

01:26:27   Yeah, well, but like, I mean, me buying new laptops is a given.

01:26:31   That's true.

01:26:32   But desktops, that's an event.

01:26:33   That's also true. I don't know. All right, we should probably get back to Ask ATP. How long until we start this?

01:26:40   Brian Hoffman writes, can you guys explain what you don't like about Apple News? Is it just the UI or is there something more?

01:26:44   Your comments over the last few years brushed it off as almost an obvious to all pile of garbage.

01:26:48   But for me, it's become my go to news source. I'm also looking to Apple One and News Plus articles.

01:26:53   Yes, the UI could be improved, but so could photos. Just wondering.

01:26:56   You know, I don't ever touch Apple News unless I end up there by accident because somebody posted a link there.

01:27:02   I don't know how to describe it, but I know that there are affordances in Apple News that are different from the standard way iOS works.

01:27:14   Like, I think swiping back actually gets you to a different article, if I recall correctly.

01:27:20   So if you do like a left to right swipe, I would expect you to go back to the prior screen.

01:27:26   But it's like, oh, no, no, here's here's another article for you.

01:27:28   What? Like, I don't want that. No, no. Every other app on my phone, when I swipe from left to right, it goes backwards.

01:27:34   And you're just sending me to a different article, which I guess is kind of backwards.

01:27:39   But that's not what I want at all.

01:27:41   Additionally, like sharing. To me, if I'm going to share a link with somebody, I want to share the canonical version of that link, which 99% of the time is the web.

01:27:51   I don't want to share frickin' Apple News article. Like, if you're going to share something, share the web version, please. And thank you.

01:27:59   I don't know. I just don't care for it. I don't like it. Half the time I end up there, it's for a paywalled article, and I don't even know it until I land there, which I guess isn't any different than the web.

01:28:08   But I just don't care for it.

01:28:10   Jon, why don't you tell us what is wrong with Apple News?

01:28:14   The second bit is what you got. The content in Apple News is content from web pages, and it should be viewed on the web.

01:28:22   The reason everyone has animosity towards Apple News is when we see an Apple News link, we're like, "Oh, I don't want to tap that and launch into a dedicated app for this special kind of URL that is worse than a web browser, that is more annoying to use than a web browser, and makes the article look worse."

01:28:37   And you know that article, in most cases, came from the web to begin with. And it's like a worse version of the article on the web.

01:28:43   Granted, the web is annoying, too. That's why we run ad blockers and everything on our devices, right?

01:28:48   But in general, the open web with URLs that are all underlink to each other as viewed by a web browser is the proper forum for web data, not a special dedicated reading application.

01:28:58   Even things like RSS readers that read a news feed from a site show a web view for the full article or launch a web browser, depending on your preference.

01:29:06   That's the right way to do this. If people were sharing links on Twitter and it was like a net newswire, you know, NNW colon slash slash, we'd be annoyed by that, too, because that's not the right way to do it.

01:29:17   If it's a link, if it's a URL, and it's a web page, take me to the web page. Don't take me to an Apple News app that embeds a web view that shows a weird version of that web page. Apple News is bad.

01:29:28   I understand people who are like, "Okay, but I like having a dedicated app for this one type of web page." Fine, good. Don't share those links with the public.

01:29:36   I can be your own little private place where you go to see things, but don't send, especially as for things like Twitter, if you want to send it to someone through messages and you know they're also a diehard Apple News user, fine, go for it.

01:29:46   But if you're writing a tweet out to the world or you're putting a link on your web page, do not make an Apple News link. That's no good. Marco?

01:29:53   Yeah, I feel very similarly. The news is almost always designed to be read in web browsers, and for all of its faults, as John was saying, for all the faults of the web, the web browser is an incredible application.

01:30:09   And we've all developed habits and expectations on how web browsers work. And you mentioned earlier, Casey, you don't like how you swipe back and instead of going logically back in the hierarchy or the history, it goes to some other article.

01:30:24   And that feels wrong, because you're looking at the content and it feels like a web browser, but worse. And if you're going to read web content, read it in a web browser. That's what it's for.

01:30:39   And if you do anything else, if you try to replicate a web browser-like experience in an app that is not a web browser, you're going to get all the details wrong.

01:30:49   And you're going to have capabilities that are wrong. Like, one of the things I love about web browsers is you can just pop up in some new tabs or you can hit Command-F and find stuff on the page real fast.

01:30:58   There's all sorts of things that are built in that you don't even think about, because they just have been in web browsers forever and we use them every day.

01:31:07   But you don't necessarily think, "Oh, I need this feature," until you're in something like Apple News and you want that feature and it's not there.

01:31:14   Or you expect it to behave a certain way and then it behaves some different way. And it just feels wrong.

01:31:20   And it has limitations and weird behaviors that just make it feel either limiting or wrong or both.

01:31:27   If you're reading web content, the web browser is the best app to do that in most of the time.

01:31:33   And anything that forces people out of that with some weird automatic behavior like Universal Links do on Apple stuff, it just feels wrong.

01:31:42   And it's frustrating and it's annoying and it can really disrupt your flow or anger you or both.

01:31:48   So I don't like things that should be in a certain format that are forcing you into their format for business reasons.

01:31:56   There's a reason why, if you look back at the sponsors of our show and you look at the sponsors of other podcasts,

01:32:05   one of the little rules that we try to enforce whenever reasonably possible is we don't like taking sponsors from services that are subscriptions of some kind of product where you pay per month and you get a box full of dog toys or whatever.

01:32:24   And it's like, "Well, I don't really need a subscription to dog toys."

01:32:29   And that feels like if you want to sell me dog toys, why don't I just order dog toys when I need dog toys? I don't actually need a subscription to that.

01:32:37   I don't like when certain things are forced into a subscription for business reasons where it doesn't really make sense.

01:32:48   And that's kind of how I see apps like Apple News.

01:32:51   This is forced into this app for Apple's business reasons, not because it's better, not because News wants to be in this weird special app that takes over this certain type of link and that kind of makes it hard to behave normally as a web browser.

01:33:08   They're not doing that because it's better for you, the customer. In some ways it might be, but in most ways it's either more limited or more annoying. They're doing that because it's better for them.

01:33:22   It serves their strategic interests or their business interests and that's why that is the way it is.

01:33:27   I don't like that kind of thing. If something is better or native to a certain way of doing things or a certain platform or a certain type of app for viewing or consuming it, I want it in that kind of app.

01:33:39   I don't want to have it in some weirdo special thing. That's why I don't listen to podcasts on YouTube. I use a podcast app because that's what podcasts want to be in and it's way better and it works the way you expect and that's it.

01:33:53   That's why every major podcast or every major audio platform is now adding podcasts for reasons and they're trying to cram it in like, "Oh, use our app to listen to this particular show that we just bought."

01:34:08   Or, "If you have this particular type of fridge, you can listen to our podcast that this podcast production company we just bought that can only play on our fridges." Nobody wants that and if you have the option of not doing that, I generally will want to take that option.

01:34:25   I like for things to be in the place, in the app, whatever it is, where they make the most sense, where they are native and where they can be full featured and follow expectations. And Apple News breaks a lot of that and that's why I don't like Apple News.

01:34:42   Kevin Duran writes, "Since M1-based computers have arrived, all tech announcements related to traditional non-system-on-a-chip computers already sound old and unexciting. It seems that on an infinite time scale, almost all computers will switch to using ARM-based systems on a chip. This has me wondering, are discrete component motherboards the internal combustion engines of computing? How dare you, sir!"

01:35:02   On an infinite time scale, you know, you get the heat death of the universe. You don't get a lot of these SECs, unfortunately. This is an interesting question because it, uh, I mean, what Kevin is pointing out is just a general trend in this part of the technology sector, which is as we are able to make smaller and smaller transistors, as the process shrinks or whatever, you need to spend those transistors on something.

01:35:28   And it turns out that if you can suddenly fit something that used to be an external chip into the, you know, the CPU or whatever, for example, putting the memory controller into the CPU, putting a GPU into the CPU, like this process of taking chips that used to be elsewhere on the motherboard and putting them into, uh, not an even bigger chip, but a chip that is more or less the same size, but now you can fit more stuff in it, has been going on for decades and will continue to go on.

01:35:53   It doesn't mean, as this process has happened, it doesn't mean that suddenly you have zero, you know, it's just a motherboard with one chip on it because you get other chips that are elsewhere, like the, you know, the cell modem or whatever. Oh, well, you can integrate that.

01:36:04   Like, so this, this process will continue, but, and it may eventually get to the point where essentially you have one chip in certain devices and then just analog stuff outside of it, if we're not already there in some scenarios, but it's not a new trend.

01:36:18   And to give an example of that, we should put this link in the show notes, an acronym from my childhood and that I also found in my schooling at various points is VLSI. Do you know without looking it up what VLSI is?

01:36:28   Oh, very large scale integration.

01:36:31   That's right. Um, and this is another, this is kind of like the, uh, high bandwidth memory. Uh, if you make up a name like this, like very large scale integration, that sounds impressive, at least they didn't say ultra or whatever, but they did say very, very large scale integration.

01:36:46   A term coined in the seventies, uh, with the idea was that you take a bunch of what used to be the street components, transistors. I don't know if you've ever seen a little black little, uh, pencil eraser thing with three metal leads coming out of it.

01:37:00   You take a bunch of what used to be a bunch of transistors on a board and you can combine all those transistors into one thing that we call an integrated circuit.

01:37:08   So all those little transistors used to be there and now they're in one thing. That's very large scale integration. They burned that term and now we can't use it for like, but just for the idea of an integrated circuit of not having separate transistors. Right?

01:37:21   So see also VHF and UHF.

01:37:24   Yeah. So, so this consolidation trend will continue as long as we are able to make smaller and smaller transistors. Eventually we won't be able to make smaller and smaller transistors and this trend will stop and who knows, maybe we'll use DNA based computing or something. I don't know.

01:37:38   There's other other avenues we can go on, but transistors will not get smaller forever because physics. Um, but for the foreseeable future, uh, we will continue to spend our transistor budget, uh, on putting more stuff into the one main chip, which we now call a system on a chip.

01:37:55   But honestly, like when, when Intel integrated the memory controller, oh, it's not a system on a chip, but we're certainly putting a lot more stuff in this chip.

01:38:02   And like, oh, we cross the threshold and you say, oh, now it's a system on a chip. Now, whatever the arbitrary definition you want to say, well, when you put this in, then it becomes a system on a chip. But yeah, we're going to see more stuff going in or the other direction you can go is, well, you can just keep making the GPUs bigger because you can basically make GPUs bigger forever and spend all your transistor and power budget on it.

01:38:19   But, um, it is advantageous to have more and more things integrated. This is another question, by the way, of like, we've been talking for a couple of years now about Apple making its own cell modems for its phones, which is a difficult thing to do.

01:38:32   And they started on it many years ago and they bought Intel's business for doing that. Right. And we fully expect if not in the next iPhone, and certainly the one after that, that Apple won't be using chips from Qualcomm for its cell modem. It will be using its own.

01:38:44   Could that be integrated into the system on a chip? Maybe, but there are analog components to cell modems that may not be easy to fold in. So maybe still be a good idea to have it outside.

01:38:57   But those are the type of things that you can imagine being sucked into that. And as for discrete components being internal combustion engines, no, because they will always be discrete components because it doesn't make sense to put literally every single thing into the system on a chip.

01:39:12   We just talked about DRAM last time because DRAM is manufactured in a slightly different way. It doesn't make sense to try to do sort of, you know, the, the, the manufacturing system to use for the logic and then a separate manufacturing system to use for the DRAM to try to put that on the same die.

01:39:28   It's not always going to make sense to do that. So there were, there were almost always be components, even if it's just like the stuff that's like all the capacitors and the analog components for the power supplies that are outside the system on a chip. That's still going to be out there. Even if it looks like it's dwindling as we pull more things into the system on a chip, we come up with new ideas for put it to put in that board space. So.

01:39:49   Thanks to our sponsors this week, Squarespace, Linode and Bombas, and thanks to our members who support us directly. You too can join, not, you know, you too, the number two, you also can join at atp.fm/join. And we'll talk to you next week.

01:40:06   Now the show is over. They didn't even mean to begin. Cause it was accidental. Accidental. Oh it was accidental. Accidental. John didn't do any research. Marco and Casey wouldn't let him. Cause it was accidental. Accidental. Oh it was accidental. Accidental.

01:40:30   And you can find the show notes at atp.fm. And if you're into Twitter, you can follow them at C-A-S-E-Y-L-I-S-S. So that's Casey Liss M-A-R-C-O-A-R-M.

01:40:48   Anti Marco Armin S-I-R-A-C USA, Syracuse. It's accidental. Accidental. They didn't mean to. Accidental. Accidental. Tech podcast. So long.

01:41:09   If you're listening to this now and you're just now, you weren't paying attention earlier, but now you're composing that email to tell me what a bread box is, just stop. I knew it when it was coming out of my mouth, Mike. I can hear, I can hear the tweets going now. You don't know what a bread box is. You've never seen a bread box. Here's a photo of the bread box on my countertop right now. Why don't you have a bread box? You're missing the point of the joke.

01:41:29   It's silly that the thing we use for 20 questions and charades or whatever to assess the size of things is a bread box, which despite you knowing what it is and despite my grandmother having one, is not the first object that springs to mind for size comparisons, but for whatever reason. Bread box got 20 questions, right? Is that the context in which you know bigger than a bread box?

01:41:53   I believe so. Yeah. Yeah. Yeah. Cause like, is it bigger than a bread box? That's, that's the, yeah. What do we want to talk about? Do we want to talk about a slack? There was a request in the chat room to talk about Tesla, which I really, really don't want to do because I can't handle the feedback. No. What did Tesla do? Do they, besides the weird tequila thing, uh, full self driving is real, man. It's real. Is it all self driving? I think it's been real like 10 times. Is the, they put a big mission to accomplish banner behind. Yeah.

01:42:22   Uh, it's real. It's real and it's in beta. Yeah. Oh great. That's something I want to beta test. Yeah, exactly. You're already, you're already beta testing and mark on bad news. Yeah. I mean, honestly I haven't kept up with this. A couple of people have mentioned this and frankly the only people have heard about it from our Tesla owners. So this hasn't even made my sphere outside of Tesla people being like, Ooh, I've got full self driving now or whatever. So I'm not even sure what's different to be honest.

01:42:51   I haven't. This is the first I've heard of this story, but I haven't been paying attention. I, the last thing I saw about full self driving is a bunch of angry Tesla owners saying that it's the worst $5,000 or whatever they ever spent in there. Super angry about it. And that made me happy because they're learning that the thing they have in their car now will not drive the car for you.

01:43:08   I love that you're saying that. Meanwhile, TKZ in the chat says it's pretty impressive. It's a limited beta. So is it good? Is it bad? Who knows? Drive your own car people until the car drives itself. That's men. And don't trust the alarm when he tells you that it does because he's been saying that for many years now.

01:43:27   Well, there's been so many asterisks on it and there still are like that every time they make progress, they do something that's very technically laudable. Like it's very impressive, but it still isn't full self driving, which is what the feature is called.

01:43:44   And like, and every time they get, you know, every time they get like one step closer to that, they call it full self driving. It's like, well, okay, can it actually drive with no interaction from here to there? No matter what I mean by here and there, no matter what it accounts, you know, what it encounters in the meantime.

01:44:02   No, it can't. Okay, well then that's not full self driving. It might be something else. It might be great, but that's what full self driving means.

01:44:10   It's not great because the closer it gets to that without reaching it, it's the uncanny valley, the more dangerous it becomes because the more it lulls the user into thinking that it really is, they really can take a nap, but they can't do not nap.

01:44:21   That's the thing like all kidding aside, like what, what level are we talking about? Because to me, if this isn't at least level four, which is defined as you can take a nap, then it's not self driving.

01:44:33   I thought that was level five. I thought level five was you can take a nap. No, this isn't level five. That's what I thought too, but then I was corrected that this is not level five. So this is very much not full self driving. This is full ish self driving.

01:44:44   But you were saying four, you were saying four is you can take a nap, right? But that's not four, is it? I don't remember. We went over this many years ago. I'm looking at it right now.

01:44:51   Level four is quote unquote mind off as level three, but no driver attention is required for safety. E.G. the driver may safely go to sleep or leave the driver's seat.

01:44:58   All right. So, so what's the difference between four and five? Steering wheel is steering wheel optional on level five.

01:45:03   I am, but, but that's that or see, that doesn't make sense to me. I guess we should have looked into these SAE levels earlier because like if you don't need to pay attention, then why does that need to have a steering wheel?

01:45:12   I mean, I suppose you could say, well, if you remove, if you can manually drive it, it's only level four.

01:45:18   Oh, here we go. Here we go. Reading my should just keep reading. That's that's the that's the trick, you guys. However, self driving is supported only limited spatial areas like being geofenced or under special circumstances.

01:45:29   So that's like it only works on the highway.

01:45:31   So here's the thing. If it says that no attention is required, that's not true. Because if you fall asleep and it ends up running, it gets off at a highway exit and then comes to a dead stop in the highway exit and you're asleep in the car because it's like, well, I'm I'm I've I've exited the area where I know how to drive.

01:45:46   So it's time for you to drive driver and you're snoring. That's not good. And it's dangerous for the car.

01:45:53   I mean, even if all the Tesla's pull over after the exit and say, well, all our drivers are asleep, but we can't drive any farther. I don't like level four level five or bust. And so far it's all been bust.

01:46:05   Yep. And I don't know, it didn't like somebody in the chat is saying, oh, well, it means it level four means you have a few minutes before you have to take over so it could like wake you up or whatever.

01:46:16   This is like the it's like the worst version of like, don't worry, you'll be needed soon, but not now. But you don't have to pay attention. But you do have to be wakeable. But like, oh, it's like how many is like trying to get kids up for school.

01:46:28   It's like, look, either you're gonna either you're gonna drive me there by yourself, or I have to be involved in the process.

01:46:35   It seems so preposterous to me and I Oh, my God, please don't email me. Please don't email any of us. I don't I don't.

01:46:41   Reminder to everyone listening that if you think that the car can drive for you, but at any moment you will be called upon to take over. You can't do that. No one can do it. No human can do that. It is against human nature. It is against the functioning of your brain. It is an incredibly dangerous situation.

01:47:00   Drive your car until it can drive for you. There is no way you can maintain that level of vigilance over a long car trip. Your attention will wander. You will not be ready to take over in time and you will die. Don't do it. Hands on hands on the wheel.

01:47:15   Agreed 1000% and please, for the love of God, please don't email us. Please, please. I don't care. I'm telling you, I don't care and I know this sounds rude. I realize that this sounds rude. I am genuinely so thankful when we get feedback because it means that you're listening. It means that you care enough to correct us or to give us new information.

01:47:37   Except the Tesla people who are all the worst. So I don't care. Don't email me. Thank you. I love you. Are you including Marco in that? He's one of those Tesla people. Not that he drives anywhere anymore. Yeah, right. I mean, but I have to acknowledge that I think of all the customer, I don't know, communities or groups that I'm in by, by like the things I buy or like, I think being a Tesla customer is the worst one of those things. Like, because like the rest of the community is so bad about fish.

01:48:05   And I know lots of people who own Teslas who are not part of the like fan base. They're like me, they own it and they like their car, but they don't participate in the community or fan base aspects of it. I mean, frankly, that's pretty much how I am with fish to like, like I, I'm not, I don't go to the shows. I'm not super into like any of the communities around it. I don't know anything about the smoke, the pot.

01:48:30   I do. Yeah, I don't do any of the drugs. I don't like, I just listen to the music and that's it. And I love the music and that's where, that's where my interaction ends. And with Tesla, like I love driving the car. It's a great car. That's where my interaction ends. Like I don't, I do not want to talk about Tesla on Twitter with people. I don't want to follow what Elon Musk is doing in his, in his side projects. Like I don't, I don't want to do any of the experimental features.

01:48:54   I've never used summon even because like, here's another thing that here's another way I could destroy my car and it achieves a benefit I don't need. So no thanks. I don't, I won't try that. Thanks. Like I just, I don't care. I just don't care. It's a great car. I love driving it. That's, that's the extent of the involvement that I want with this brand.

01:49:15   And that's what's like the relationship I have like with the brand of scissors that I buy. Well, you know what I, when I buy, when I need scissors, like, you know what I do? I buy some scissors and I don't know a damn thing about the brand. I just know I like these scissors and I buy them and that's it. Like that's how I am about my car. I like this car. It's a great car.

01:49:33   I don't care at all about the company. Don't want to argue about it. Don't want to talk about all their weird, you know, side projects they're doing and their beta features that almost kind of sort of fully work like nope, don't, don't care about that. I just like the car. It's a great car. I like driving it.

01:49:49   [beeping]

01:49:51   (beep)