"Company developing Mac Pro desktop with 40 cores, new Mac mini"
It will be interesting to see how close the rumors are to fact when real hardware announcements happen. Given how fast the M1 in the MacBook Air is performing for tasks like video editing I wonder if they will really need to push to such extremes as putting 40 cores in a Mac Pro -- unless they plan to edit four 8K streams at once?
The key part there being "scientific research server." The Amazon Graviton2 ARM chips have something like 64 cores but that makes sense for machines that are powering AWS. Unless the point of these Mac pros is to be at supercomputer speeds or to pack massive specs like VM packs a Bughatti Veyron (because we can) I don't see much of a point beyond the highest of high end video editing and scientific research.
I mean Mac Pros can easily be configured to $50,000+ machines from their base price of $6,000. They aren't aimed at the average joe. The majority of these things are being sold for business purposes in the first place with maybe a handful of sales to ultra rich enthusiasts.
The 40 core model is probably the replacement of the 28 core Xeon which is a $7,000 option.
If you're going the scientific computing / HPC server route, then the sky's the limit for core count, you can buy dozens of these and connect them up with Infiniband to run your MPI job or whatever. The point is, the people buying Mac Pros are usually content creators (photography, video editing, music composition, animation etc...) rather than researchers for numerical computation, and for them Mac Pros are probably going to be sufficient enough (unless you are doing 3D animation and need render farms). Most number-crunching scientific code is written only for Linux, you'll never buy a Mac anyway.
There are plenty of editors who would love to be able to edit 4 8K (or 12K!) streams in real time :)
I still deal with 1080p and get annoyed at my hardware's limitations. If I could justify a Mac Pro and Apple's accelerator card, I'd get them in a heartbeat.
For certain applications, no matter how good the hardware you have is, there's always a desire for more, faster hardware.
"There are plenty of editors who would love to be able to edit 4 8K (or 12K!) streams in real time :)"
Is that really a common scenario? I can see editors at the big studios using monsters like this (because the studios can afford it) but are we talking about a market of hundreds or thousands? Is this something a free-lance video person would use?
No, 8K is not really reality at the moment, but editing UltraHD is reality, and being able to easily edit or composite 2/3 streams of UltraHD with some complex effects in realtime is desirable.
It's a bit mind-bending for those of us who remember the SGI Octane and Onyx days to think that 100x as much computation is available at 100x less power-consumption. (That's 10,000 times more efficient, for those following along at home.)
Between that and the almost infinite instantaneous computation power available in the cloud at the press of a few buttons (and sufficient capacity on one's credit card), we ain't in Kansas any more.
> Apple last fall started replacing Intel processors in the iPhone and iPad with the house-designed M1 chips, which eat less power and let the mobile devices run the same apps as Macs.
Really makes you wonder what other statements there are outright false. Prior to Apple's own ARM chips iPhones used Samsung processors.
Interestingly though, Intel did at once point make very nice mobile CPUs (notably the XScale PXA270, which clocked in at a little over 500 MHz and had a halfway decent PowerVR graphics accelerator).
I mean, this is coming from Bloomberg. Y'know, the outlet that still hasn't retracted, apologized for, or in any way acknowledge the fact that their "Big Hack" story was a giant steaming heap of bullshit.
Isn't part of it that there's basically no bezel on the 14 and 16 models? I know Apple minimized that over the years it really is essentially zero now. I haven't checked but I wouldn't be surprised if the footprint of a 16" MBP is actually smaller than most 15.6" (or are they 15.4"?) non-Apple laptops.
I think since they've managed to minimize the screen bezel, they can fit a 14" screen in the body of an older 13" and a 16" in the body of an older 15"?
Yes, this is the case. (As the owner of a 16" MBP, which replaced a long series of 15" MBPs—with a 17" in there way back when they were still current.)
They're selling the same two size laptops, just with larger screens due to smaller bezels.
Right now I'm looking at a 2019 16" and 2018 15" models, they're pretty much the same size, the 16" model is just a couple mm wider. Probably the same applies for 13"->14".
For Windows support, I thought Apple said something along the lines of: "If Windows wants to add support, we are fine with it" - but I could be mistaken.
Native Windows on M1 would be interesting. But it could potentially embarrass all of Microsoft's hardware partners as well as Microsoft itself (cough, Windows On ARM). Also the market for it would probably be measured in whole handfuls. So I don't think we should hold our breath.
So they are releasing a new MacMini within one year? That is pretty amazing. Previously they would go at least 2 years between revamps. I guess the MacMini is becoming more important?
They still sell an Intel Mac mini with more RAM and better I/O. This wouldn't be a revamp of the current M1 mini but replace the Intel Mac mini as a high end model.
I'd argue it has much less to do with mac mini being any more important than the chip itself being important. They're probably trying to push it in every spot they can, and there is no sense in making more old chips if you can just put the new one in instead.
In a way it's possibly more efficient without the heat pipe - the cooler is directly over the CPU in this design, and can be quite large due to the way it's packaged (though it has to be thin).
The downside is it's harder to get decent airflow over the fins. The casing is probably designed to direct the airflow, but I'd have to see all parts together to figure out how this works.
Correct. It needs to be completely sealed. You can't take the back off the laptop and have the cooling system function, the case is part of the cooling ducts.
Provides for heat dissipation over broader surfaces. This is normal.
Because it helps gets more views on YouTube which is the only thing he is after. Him patronizing the thermal design of Apple computers that way is pretty funny, ngl.
Idk, I am no expert but I tend to believe that a design that is shipped to millions of customers and only a handful report problems is probably designed better than what the average youtuber or forum commenter can assess. Apple is known to remove elements from their design in order to optimize. That big and heavy heat pipe seems like a perfect element to remove.
And maybe it is a mistake. That happened before and they iterated over it and turned the design into an advantage most of the times.
Judging and say "does nothing" is very naive in my opinion and I am pretty confident that everything in that designed is planned pretty accurately.
The M1 MacBook Air has no fan, so you can’t really compare its cooling design to the previous MacBook Air. Given that it’s fanless, it would hardly be surprising if it had bigger passive cooling devices.
The idea that a cooling fan needs to “blow on” the thing that it’s cooling is just dumb. As long as the path from the fan to the CPU is sealed, it makes no difference at all how far away the fan is. (Think of a wing being tested in a wind tunnel. The fan that’s making the ‘wind’ doesn’t need to be placed right up against the wing for this to work!)
A heat pipe gives you more layout flexibility because it lets you place the fan somewhere where it can’t generate airflow directly over the CPU. Given that a heat pipe can’t be 100% efficient at transferring heat, a design with a heat pipe is actually less efficient, all else being equal.
Only if the input and the output of the fan are piped together.
Apple was designing stuff that considered the airflow THROUGH THE CASE back with the first Macintosh, the 128K. You could even get a silly looking chimney that used convection to pull even more air through the case with no other modifications.
All Macs with air cooling, not just the laptops, are designed with internal airflow in mind and always have been. You do not have a little fan blowing directly on the CPU (from where? to where?) and expect it to do anything.
I don't know who this fellow is, but next he can take the top off his cylinder head and then be upset the engine doesn't work. From first appearances he is an absolute bozo.
He's pretty much the main figurehead of the right to repair movement from the consumer electronics side of things. He has run his own independent repair shop doing board level repairs for years in NYC.
Then from second appearances he has other things to recommend him. As far as laptop cooling, if he isn't joking or trolling about expecting the laptop cooling to work with half the ducting removed, he's a bozo about that.
That’s not a valid criticism of the cooling design of that computer. It may or may not be an adequate design, but this video doesn’t really address it.
There’s no particular reason the fan should be physically next to the CPU or heat sink. Its job is to draw air into the computer, across the heat sink, and then back out.
You generally don’t want the air intake and outflow to be too close to each other since that would typically cause heated air to be drawn back in, making the cooling a lot less effective. And you mostly want the air flowing in a channel (rather than blown all around inside), so you can expel most of the heated air.
So, e.g., you have intake on one side of the computer, a channel for the air to flow inside the computer, and a fan at the other side of the channel, pulling air through.
This will be effective if the heat sink on the CPU is anywhere along the channel. There’s no reason it needs to be next to the fan.
A piece of copper, BTW, is a way to move heat from some place that doesn’t have good exposure to the air channel to a place where it does. You only need to do that if the heat sink doesn’t already have good exposure.
To do this properly, I think you’d want to analyze the design and then run some tests to see under what circumstances heat builds up.
Lost a little respect for that guy. He’s a smart guy, so I know he understands this stuff. You’ve got to wonder why he’s bullshitting his audience like this. Seems clear to me this video is about promoting himself rather than discovering or explaining flaws in the cooling system of Apple laptops. (Or maybe I’m giving him too much credit, and he really doesn’t understand this stuff. Either way, it’s not a good look for him.)
The Mac mini used to have a 65 W Intel processor. If the M1 uses ~15 W then they could fit something like 16 big cores, a few small, 32 GPU cores, maybe a 256-bit memory bus, etc. in the existing case.
> Apple plans to launch the redesigned MacBook Pros in 14-inch and 16-inch screen sizes. They’ll have a redesigned chassis, magnetic MagSafe charger and more ports for connecting external drives and devices. Apple is also bringing back the HDMI port and SD card slot, which it nixed in previous versions, sparking criticism from photographers and the like.
Even if they continue to force the anti-feature that is the Touch Bar onto users, I'd immediately buy an M1 MBP with MagSafe and actual ports.
I've had MacBooks since 2005, and I'm an amateur photographer. I really don't miss the ports or MagSafe at all. I actually like the USB charging better, as it allows you to connect the cord to either side, and it doesn't get accidentally knocked out if I use it on my lap. I've found that the cord being on other side is nearly as good as magsafe anyway, as it will be better positioned to pull out if it gets yanked.
Yes, I did use the SD card reader from time to time, but most serious cameras use other cards (my older camera uses compact flash, and the newer one uses XQD). It makes no sense for Apple to integrate readers for those, and my XQD/CF readers also read SD cards, so the card reader is useless.
All of my mobile peripherals at this point are USB-C, so I don't care about USB-A ports either. And when I'm at home, I use a dock, which has all of the ports I need.
Honestly I'd much rather that they keep the laptops thin instead of adding ports I'm not going to use.
The vast majority of cameras, even professional mirrorless or DSLRs, currently use SD cards now. Perhaps that's changing, but I would imagine there's sufficient utility to the SD card reader that many users would value that feature.
That's not true for professional mirrorless/DSLRs, which on the high end often use other cards like CFast, XQD, or CF. And as mentioned, many readers for those cards also integrate SD readers.
The vast majority of people these days just use their phones for cameras, and don't use point-and-shoots or other things with full-size SD cards.
So really SD cards cater to just the mid-range camera market, where people use mirrorless/DSLRs but not high-end versions that often use faster cards.
I don't think it's a good idea for Apple to make everyone's laptops bulkier so a small minority of people can avoid using a $17 adapter like this:
For a company that prunes as ruthlessly as Apple seems to, including an SD slot in this day and age seems like an odd choice. I'm even a photographer and I rarely use the SD card slot on my MacBook Pro and certainly wouldn't consider tossing a small/cheap adapter into my travel electronics kit a big deal.
As someone else mentioned many of the higher end new cameras are moving XQD or something similar that is faster I presume to deal with video.
With that said, I have cameras that use SD cards and don't see the need for an SD port. I have usbc adapters that are not much bigger than the SD card that I can use on my MBA or iPad. With LR, I can basically load my days pics on either device and have them available everywhere.
I'm very skeptical they are adding back SD, so we'll see. I think MagSafe is going to be something like what came to the latest iPhone and not what people think of traditional MagSafe. Or maybe they'll add something to the usbc port that works like MagSafe if using an Apple cable.
USB-C would have been a great replacement... for DisplayPort/Thunderbolt ports.
It's finally, years later, getting merely not infuriating to be stuck with only USB-C ports.
Single port for display + charging is very nice, when it works.
That iPhones and most iPads still don't use it is annoying. Having two kinds of charging cable around, years after Apple went all-in on USB-C for laptops, sucks.
[EDIT] two kinds of charging cable and still needing to buy & carry two kinds of adapters/dongles for everything. Ugh.
Suspect at this point Apple doesn’t want to break compatibility with their Lightning ecosystem when wireless charging is probably going to be the future on iPhones.
The most recent non-pro iPad does have USB-C and I expect whenever they get around to redesigning the lowest cost iPads they’ll also move to USB-C.
Apple uses lightning on their wireless keyboards and mice for chargring... I mean, what the fuck! They even came with lightning to USB-A cables, which won't plug into any USB-C Mac. Also it didn't come with a power brick presumably "to save the environment." I literally had no way to charge them. Totally asinine.
I am pretty sure they come with Lightning to USB-C now. At least the ones included with the new iMac do, and iPhones do so it would make sense for them all to.
> and it doesn't get accidentally knocked out if I use it on my lap
Man, my 2017 MacBook Pro's USB-C charging cable (the official one, not third party, and in pristine condition) constantly disconnects itself to the point where I'm basically traumatized from the annoyance of it now. It only provides power if it's ALL the way in, and the damned thing just pushes itself out half a millimeter and cuts power repeatedly.
I really want to like the generic nature of USB-C charging, but proper magsafe with a strong magnet is just miles better (third party magnet adapters for USB-C both break the spec and also tend to have very weak magnets which can cause damage from sparking at the port).
There's nothing to clean. Everything is pristine. The problem has even persisted across replacement cables and a full top case replacement (the keyboard, of course) where Apple scrubs the whole thing down before they give it back. Maybe I should superstitiously blow on it like it's an NES cartridge. It won't help, but maybe I'll feel like I have some agency.
I had similar problem with MagSafe; the culprit was the dust that both the charger and the laptop collected over the years. After some compressed air treatment, it was as good as new.
I have a 2017 for testing at work. It has never moved. I ssh into it for some macos specific testing. I have plugged and unplugged the charger maybe 5 times.
Yesterday I moved it around. And noticed what you described. The plug will come back out by a millimeter. And then it won't charge.
Man, that half a millimetre difference between connected and disconnected drove me nuts. It is absolutely infuriating because there is zero physical feedback when you try to connect the cable.
I don't have anything helpful to offer but I feel your pain. Reading your comment almost brings on PTSD from remembering what it was like.
I actually restored my 2015 Macbook Pro and have been so happy and content using that in the past two years or so. Not sure about the other sizes, but my 13" from 2015 does everything better than the one from 2017. Performance, keyboard, connectors. None of the overheating bullshit where the fan is constantly going and performance is throttled.
FWIW, I was very unsatisfied with my 2016 MBP USB-C ports and found them very unreliable, but am quite content with the ports on the mid 2019 MBP and the 2020 MBA (M1). So, maybe they've improved over time.
> I actually like the USB charging better, as it allows you to connect the cord to either side, and it doesn't get accidentally knocked out if I use it on my lap.
I felt the same until my cat pulled my laptop off my desk by the cord. It landed on the corner with the USB plugged in, destroying the port on the laptop itself. Probably knocked a few hundred off the resale value. The little shit.
Maybe the difference is that I'm not often using the laptop with the cord strung out around animals (not that there is anything wrong with doing that). For me, at home where I have a dog and several small children, the laptop is usually on the dock. The same is true at work. When on the go, I don't usually have the dog/small children.
Apple’s goal is to have every portable computer work like an iPhone: you charge it overnight, then unplug it and use it unwired for a full day. This is why the charging port on the mouse is on the bottom, for instance.
They never quite got their Intel laptops there, though. (Especially running MS Teams, Slack, and Chrome!)
It’s kind of funny that, now that they can make a laptop that really does last all day on one charge (according to tests of M1), they’re switching back to MagSafe.
Unfortunately that doesn't work for those of us that need to run compute intensive tasks, like browsing the web, or compiling stuff, or running docker.
There's this stupid dream that the computer needs to be light and fanless and thin and sexy. It doesn't need to do that. I'm happy if it doesn't melt while trying to run some software I need to do my job.
My experience with my mpb M1 16GB is that it barely every spins up the fans, even when pegging 2 cores to 100% because web.whatsapp is doing it's thing again and I have a Python process doing some hammering. Docker in silent model is just consuming RAM but not really consuming battery. The only thing i've noticed that is really heavy is jitsi. But Discord or Teams are pretty chill on the battery life to my experience
One thing Lenovo gets right is that they have fairly clear desktop replacement and travel laptop personas. I'm pretty clearly in the latter category although I'm hoping that the new rumored 14" will be able to replace both my 2015 MacBook Pro and iMac.
> run compute intensive tasks, like browsing the web
lol, it's not only browsing though but web-infested Electron apps as well. Used draw.io (which is otherwise a fine diagramming app) the other day, with fans spinning up all the time. It's sad to see that draw.io was (and internally largely remains) originally a Java (Swing) app, seemingly only ported to a fscking browser frontend as a fashion statement when the exact same things it does I could do almost 20 years ago on a PowerBook using OmniGraffle.
Each to their own but magsafe saves my laptop frequently, it’s nice to not have to worry about it being yanked off a table. I have a ton of usb a stuff that doesnt have a usb c counterpart, and likely never will. I tried to use a USB drive to move stuff between two computers and we ran into a “one is usb a only one is USB c only” problem
Regardless of the potential merits of alternative storage formats, it's a good point that people who have needed an SD card reader since their removal have probably gone out and bought external ones by now. I'd expect the standalone units to generally be higher-quality than the ones that would be built in.
That said, I'm glad to see Apple has finally realized some of the ways they've been alienating their customer base, and they appear to be walking back some of their most unpopular changes.
Apple has sold their dual GPU late 2015 powerhouse model on the website, allowing people to completely skip the USBC donglebook with butterfly keyboards and touch bar
I’m looking forward to having that in the new processor and with touchbar
They are acknowledging the folly of their ways as they have gotten rid of the butterfly keyboard, adding magsafe back, and adding sd cards back
They are making the least contentious and perfect machine
As a musician with tons of peripherals, some of which are annoyingly sensitive and can't be used through hubs or anything but straight-through adapters (my turntable mixer's digital interface, for example), I welcome the return of the ports. Its not uncommon to have an audio interface, one or more midi controllers, the aforementioned mixer all connected for instance. And basically all of those things still use USB-B ports for whatever reason, and I never see USB-B to C cables ever.
The last time I searched amazon for usb-b to usb-c cables I genuinely couldn't find any. I bet I would have better luck searching for 'usb-c printer cable' as someone else pointed out, but I didn't think of that.
Last time I needed a cable on short notice before a gig I had to settle for an adapter, and was really lucky that it worked. I've had other seemingly simple pin-to-pin adapters not work on me before.
Next time I'm in a jam I'll run down to my local ebay though, thank you for this super helpful reply.
Just bought an USB-C to USB-B cable for my Macbook Air m1 and an iConnectivity midi interface and it didn't work.
(Brand was Equip). Another cable from Delock to mini USB-A was fine though. Some of these cables seem to be only for charging so do some research.
Best bet seems to be Apple adapter even tough I had them break every year or so.
My impression from these reports is that Apple is planning to differentiate their laptops more strongly than they have been: there will be ones that place "thin and light" as a priority, and ones that place "give me more I/O, dammit" as a priority. If you want a high-performing Mac laptop that's thin and light and you don't care abort ports (hi, this is me), you can already get an M1 Air that has the same performance as the M1 MBP (and this is exactly what I did). The M1 Air is good enough that you're not really losing anything other than the perceived cachet of putting "Pro" on the nameplate. (Well, you are also losing the Touch Bar, but one can argue that's actually a net gain.)
From these reports, the next revision to the Air is going to get a bit of a redesign, colors, faster CPU, etc., but "Air" is still almost certainly going to mean the line that prioritizes being thin and light. You will still have a Mac laptop that gives you what you want. But people who do want more ports will, for the first time since Mac laptops went USB-C, also get what they want. (Or at least get closer to it.)
There’s SD cards for 99% of all hardware, and then there’s a ton of different standards for all these niches. Like CompactFlash, XQD is a dead format already. Apple isn’t going to add ports for that.
You hate the Touch Bar so much that if Apple's next round of laptops includes one you'll stop using macOS and start using Linux (or Windows) on your laptop instead?
Sadly not that much, but I for one really dislike their decision on keeping it. It adds no value at all to my workflow; quite the contrary - It distracts and disrupts me a lot. Would love to get rid of it; but I learned to ignore it. There are apps that can completely disable it (Nocturnal).
Right now I have an Air. Besides price, the main reason for getting an Air was not having the touch bar. That's non negotiable for me.
I want to both beef up the memory/processing power of my laptop and switch to Linux (which comes with built-in performance bonus of native Docker). So I am headed to ThinkPad anyway, but there's a chance Apple could keep me locked in if they make a nicely-powered proper dev machine with no damn touch bar.
Migrating to Big Sur has been a pain in the ass even though I am still on an Intel model.
Tons of Python language versions and packages are a total PITA to get built on Big Sur.
To be fair this seems to be a pretty Python-specific issue, but the inability to do much about it really has dug the feeling into me that I don't own/control my machine, Apple does.
They released a new keyboard with TouchID and no TouchBar. IMO, that means the TB is dead. Otherwise, it's a niche item that no one will ever make full use of.
I wouldn't take that as a clear sign. They still made an M1 touch bar 13" MBP, they bothered to refresh it and keep it when swapping from butterfly back to scissor and the MBA only has touchID too.
The touch bar is handled by the T1/T2/M1 chips in the laptops. It could be that just cramming a T2+OLED into a keyboard may have made it way too expensive for a keyboard. Plus does Bluetooth even have enough bandwidth to handle sending screen data that size reliably?
It is a niche item and it does suck but for some reason they keep tacking it on...
> They still made an M1 touch bar 13" MBP, they bothered to refresh it and keep it when swapping from butterfly back to scissor and the MBA only has touchID too.
That's because changing the chassis is too expensive, but when they redesign it, you'll see it most likely without a touchbar.
Not exactly sure if I'll end up buying a MBP later though. I basically only use my Mac for recording audio, and once you get your mic set up , your not too portable anyway
Is Apple even going to refresh any of its Mac products with Intel CPUs at this point?
When they first announced the transition last year, it was implied that some products will still get refreshed during the 2-year transition period. But based on this article and common sense, there is minimal incentive to do an Intel refresh. Refreshing both discourages buying of Apple silicon-powered Macs and lessens the performance jump that Apple is so eager to publicize.
For example, Tiger Lake H-series was just announced/released and it'd logically target the Macbook Pro 16.
I'd be a little surprised if Apple release _any_ new Intel machine. I'd be _amazed_ if they release an Intel laptop. It would be generally slower than the 13" M1 (Anandtech has Tiger Lake H _just_ beating the M1 in for multithreaded workloads in short synthetic tests, but given the throttling shown (https://www.anandtech.com/show/16680/tiger-lake-h-performanc...) even in a presumably friendly reference laptop it wouldn't be competitive in the real world) and have far worse battery life, and be replaced in under a year; what would be the point?
They refreshed the iMac with new Intel processors two months after the Apple Silicon transition was announced, I think that is what they were referring to. I doubt any other computers will get new Intel processors at this point.
Large integrators like Apple usually have a tie line to inside Intel (in this case, probably at the CEO/VP levels) that would feed information to Apple regarding product lines.
Yes, Tiger Lake H-series was just announced, but it's likely Intel proposed selling this chip to Apple 2.5 years ago, and Apple didn't like the roadmap, and said No Thanks.
That's how much time supply chain changes like the move to apple silicon chip will need - I wouldn't be surprised if Apple already knew of all of Intel's roadmaps 5 years into the future, and decided, nope. No thanks. Not giving any more money to Intel.
From this point of view, you can tell, it's detrimental to Apple to go back to Intel now, if they wanted to use the Tiger Lake H they would have used the threat of M1 as a bargaining chip to lower the pricing on upcoming chips. It's obvious they didn't do that.
This chip is going to give us surprises in the next future. I can only imagine what is is store, but a rack of this kind of tech could really drive down power consumption of data centers
It might not be a 19" rackmount chassis like the Xserves of yore, but it's hard to deny Apple hardware will continue to find its way into real datacenters.
The high end Mac mini SKU mentioned in TFA sounds great for this application, especially now that they have 10Gig Ethernet.
The point is driving down the power consumption in datacenters. I doubt the Apple hardware at AWS is cause of high power consumption, and I also doubt Apple hardware will replace Linux and Windows machines that compose the vast majority of AWS computing power.
ARM might replace Intel, but it won't be branded Apple, that's for sure.
I'd be a bit concerned to write something with parallel processing, only to have an overheating issue denied by Apple for years. Butterfly keyboard flashbacks...
It’s definitely not easy. In general it is very hard. Some applications have been optimized and for a few applications the fit is very natural (ie, where it is easy to avoid single threaded bottlenecks, or where you don’t need to pass data from core to core via the cache). But for general purpose workloads, it will take a lot of optimization.
Agree to disagree. There's a large host of applications that can benefit from parallel processing power (think, anything that can take advantage of a GPGPU).
I don't dispute that some applications are hard to parallelize. However, in my day to day work, I've found that parallel operations are often not that complicated to introduce.
The bigger issue is that CPUs are now so fast I find myself blocked by IO far more often than I'm blocked by actual processing.
For general purpose workloads maybe, but the target-audience for the Mac Pro is really for people who are using applications that are highly multi-threaded. Just look at the use cases Apple describes the existing Mac Pro for:
> From production rendering to playing hundreds of virtual instruments to simulating an iOS app on multiple devices at once, it’s exceedingly capable.
> Increased speed for Logic Pro X, MATLAB, Adobe Photoshop 2020, Autodesk Maya, Wolfram Mathematica Build Time, Final Cut Pro X
> Tasks like animating 3D film assets, compositing 8K scenes and building lifelike gaming environments.
This is the target market - and pretty much everything listed here tends to be multi-threaded. Making multi threaded applications aren't easy, but for high performance tasks all these use-cases have to be multi-threaded where possible or they are just too slow.
I think that people grossly underestimate the amount of processing that can be done on video. Reasonably sized filters, something like deinterlacing, high quality transforms. Swamping a modern CPU in terms of memory access or processing isn't too hard.
Sure, it's likely why AMD 64 core Threadrippers are overtaking the pro video industry at about 1/3 the cost of Apple's current top of the line Mac Pro. 8k/12k uncompressed RAW video is very demanding. It will be interesting to see how this 40 core apple cpu competes with the 64 core. Maybe they'll get a bit of market share back, but it better outperform by more than a few percentage points especially if they will cost more, which is likely.
Not quite that simple: Using AOM AV1, a 12-core Ryzen 5900X beats a 128-core dual EPYC 7763 (same Zen3 architecture) in every encoding benchmark, according to https://openbenchmarking.org/
You have to be smart about how you process the video.
Yes, if you just use AOM directly you'll have a hard time saturating the core with work.
However, if you split your video up into scenes and start an AOM instance per scene, it becomes trivial to outperform the 12 core ryzen with a 128 core machine. The main bottleneck becomes memory bandwidth.
All that said, you need videos long enough for there to be a benefit here. The Ryzen will still win if you are talking about making 10 second av1 gifs.
Rust makes writing multithreaded code stupid-easy. The compiler itself will saturate every core of your machine. Rust development on a threadripper is such a joy for this reason, and even the M1 is quite pleasant, so I'm looking forward to these higher core counts.
The other article was speculation about a new MacBook Pro (laptop, possibly this year). This is speculation about a new Mac Pro (desktop, possibly next year).
The source says that the new chip for the high-end Macbook Pro would have 8 performance cores and 2 efficiency ones. It's a bit surprising that they'll reduce the efficiency core count to half (although this might get offsetted by the single-core advancements in the new chip)
As someone who just invested in a Mac Pro late last year, I'm a bit frustrated. I wanted years for a modular design that I could keep and upgrade over a longer time only to find out I'll be forcefully phased out within 2 years.
M1 mac pros were always going to come. Apple themselves announced that the next 2 years will be transition.
I'd say if you purchased a Mac Pro, it's still the same. If you find yourself not having enough power, you bought the wrong machine to begin with (probably modern threadripper pro would be what you should have been looking for)
If the computer satisfies all your performance needs, what use is it to "feel frustruated" over something that was expected to come?
Don't feel Fear of Missing Out. The Mac Pro didn't change. Your feelings about it changed because of marketing and advertising (ie apple trying to persuade you that you need the M1, when in fact your computing needs are being taken care of already)
Wait for the next generation after that current M1/M1x mac pros and go for the next generations when apple comes out with 128 Cores and 256G of ram.
One constant of computer industry, there will always be the next chip, the next computer, and by then, the fancy M1 Mac Pro will become obsolete. Exactly like how nobody wants a 486 computer anymore but when it was released it was the hottest thing on the planet. Its day will come when everyone says "oh that M1? that was a slow POS that is now ewaste"
"what use is it to "feel frustruated" over something that was expected to come"
1) The details were all speculation.
2) There was no other option if you needed more horsepower.
3)_I'm happy with the current Mac Pro, but its going to be phased out of software support before its time, just like the Motorolla.
So it's more about the software support being phased out than new hardware coming out.
1) it is not speculation that there will be an M1 Mac Pro. Apple clearly says within the next 2 years there will be no more Intel macs for sale.
If you want the software support, wait until the support dies/runs out, and then buy the newest computer at that time, when you need it.
You'll find that apple will drop support for the M1 too in the future. Its day will come when Apple decides the neural engine is too outdated or the architecture is too old and will stop support it as well. That's a fact, not speculation.
> ...but its going to be phased out of software support before its time, just like the Motorolla.
This is precisely why I as a consumer do not buy Apple products. Apple has a long history of dropping product support with no upgradeable path forward, only buying new hardware.
This is also, in my speculation, the reason Apple is one of the most valuable companies on the planet, even though they are primarily a consumer products company. There's a constant internal urge for the newest shiniest product (which consumers of all products feel), but also the very real external requirement to buy new if you want to work with anything new.
My best anecdotal experience for this was when my uncle contacted me because his wife bought him an iPod Touch, but his Mac was telling him he needed a newer version of iTunes (I believe he bought the Mac 4 years prior). It turned out however that the new version of iTunes couldn't be installed on his "outdated" Mac because it was bundled with an OS update that his machine was too old to receive. I meanwhile was able to load iTunes under Windows on a machine twice the age of his Mac and connect the Touch without issue. He didn't want a new Mac, didn't need a new Mac, but chose to buy a new one to work with the rest of his ecosystem.
Apple typically has the last newer hardware support of any phone or pc manufacturer. My daughter still uses a 2011 iMac without problems and with relatively recent security updates.
>(probably modern threadripper pro would be what you should have been looking for)
Comments like this are always just so empty to me. This person was looking for a more powerful Mac, so suggesting they should have been looking for a CPU that's never going to be available for use in a Mac is just grandstanding. This person had been waiting for years for a product from Apple to replace the worthless trashcan systems. How your suggesting of something never to be available is helpful eludes me.
Of course, as soon as the M1's were announced with the 2 year transition plan, the actual suggestion would have been to wait until the M1 mac pro comes out.
My suggestion was if there wasn't enough power, and the person needed a computer (which is the only reason one would buy a mac exactly during this changeover point), then a threadripper pro running linux would be the way to go for many workloads.
The alternative is to just wait a few months and see what comes.
> if you purchased a Mac Pro, it's still the same. If you find yourself not having enough power, you bought the wrong machine to begin with
This is the reality really, if you actually wanted power then the Mac Pro just wasn't what you should have bought. Threadripper + Nvidia cards would take you way further for far less and with a clear sustainable upgrade path.
Also really after the past 10 years of Apple's pro moves with their last 2 "workstation" machines being discontinued it seems very risking investing so much money in their pro line.
Yup exactly everyone bought the trashcan mac (I have one now myself - I love these buggers!) and then were disappointed with the lack of expansion/improvement.
Everyone then bought the current Mac Pro, but were disappointed that nothing else really fits into the slots aside from the things that Apple offered themselves (Nvidia cards won't work in the current Gen Intel Mac Pros.)
What makes anyone think the next Apple Silicon based Mac Pro will be any different?
Wanting a more modular mac pro is more like wishful thinking or wanting something to be different when reality is clearly saying something.
I'm currently grappling with an issue in my Linux VST development, where specifically Threadripper users are having issues. This is on Github and there are better people than me tackling it: it's an open source project so people can try everything they want.
So far the answer seems to be, there is no getting around these CPU issues through tailoring the compiler flags etc. and the best people can get is amelioration of the problem, not getting it back on par with normally inferior Intel chips. Again, the issue is that the Threadripper stuff is eating unreasonable CPU and the Intel stuff, normally not as good, is not doing that.
I suspect what you're looking at there is a CPU architecture that's getting optimized for specific conditions, which my plugins don't meet. There seems no fix other than 'you should be playing Mass Effect on this computer instead'.
I'm sure that it would destroy the Intel computer if you're doing that, but generalizing that to all conditions seems not justified.
When the "new" Mac Pro was announced, the shift to ARM processors for MacOS was at most rumored and it also wasn't obvious that such a CPU would perform well.
If people knew this was "going to come", they would likely not have invested a significant amount of cash into a platform that was already on its way out.
> One constant of computer industry, there will always be the next chip, the next computer, and by then, the fancy M1 Mac Pro will become obsolete.
Unlike the incremental improvements you get with each new Intel CPU generation, this is quite a disruption. The PPC Macs soon were obsolete after the Intel Macs came about. All the efforts went into the new platform.
> Its day will come when everyone says "oh that M1? that was a slow POS that is now ewaste"
Things don't move that fast anymore. Take a 10-year-old Intel CPU, it's still quite capable. That certainly wasn't the case 10 years ago, or 20 years ago.
The Intel MacBook was introduced in 2006 and SnowLeopard discontinued PowerPC in 2009. So 3 years seems to be about what you can expect before there's no OS, and another 2-3 years after that the patches stop.
At some point macOS will no longer provide an Intel version but that is likely years down the road - perhaps 2026 at the earliest? And likely much later.
And third-parties will supply upgrades as the upgradable parts are roughly standard, so the Mac Pro is likely to be a completely usable machine for nearly a decade to come.
They’re not. The Mac Pro modular design has always been a lie. Yes, it’s modular, but there’s nothing you can put in there because there are no useful drivers.
I got an iMac Pro. I have similar feelings, but for all that it's certainly doing the heavy lifting I need right NOW.
Feels like wanting to be both on the cutting edge of performance, and running/developing for a Mac, is damn pricey. However, I'd rather have it this way, than sit with my current machine being the best you could get. In some ways I am gratified that Apple is busy making last year's news feel really inadequate.
You can be damn sure I'm devoting a lot of effort to promptly supporting the new architecture. I've already got literally hundreds of M1 builds for my Audio Unit plugins out. I chose to produce native versions of EVERY product, no matter how old it was: rather than make a new product and only release that for M1, I treated it as the same as the 32/64 bit transition, and re-did every product in the whole library.
If we're going to get absurdly capable Mac computers, I mean to be categorically included in that equation.
For now, I'll enjoy my very costly iMac Pro that's still very capable… AND the M1 Macbook Pro at a fifth of the price that outperforms the iMac Pro at a few things already.
It seems obvious once you look at it - the "efficiency" mode is almost a quasi-sleep and so it makes sense to only dedicate a certain number of processors to it (as every efficiency core could in theory have been a performance core).
The OS appears to partition the processors, like linux cpuset stuff, with low priority threads getting given to the efficiency cores, so you basically keep the low priority background stuff off the performant cores.
How many is optimum is an interesting question, I would assume the apple peeps have a good idea and are on the money with this design.
it's not clear to me why the optimal number of efficiency cores wouldn't be zero on a desktop. why not just use the die area for additional performance cores? one or two full cores shouldn't have much effect on the thermal/power budget if they're just working on background tasks.
Perhaps. It wasn't obvious to me. I thought things like non-active browser tabs and various background daemons would easily occupy more than 4 efficiency cores. Though I suppose that has a lot to do with how they are implemented.
I assume they optimize Safari for this, and I'd think they'd be much more likely to quiesce a background tab (and let it swap out) than to keep it running at a slower speed.
At some point it's more efficient to have a single performance core going than a number of efficiency ones.
They do up to 8 on the big CPU according to rumor, so there's obviously some room there.
Consider the use cases. The Mac Pro is a desktop, always plugged in and used for the power. The M1 is on a laptop running on a battery where efficiency is more important to improve the battery life.
A desktop runs a lot of low priority processes. The efficiency cores can be given all those processes. This means some low priority thread will get scheduled on a performance core and bust the cache.
There's also plenty of time where the system has low intensity use or is idle. The efficiency cores let things like receiving notifications or checking e-mail happen for very little power draw. Fans only need to run during intense use and the system can be quiet the rest of the time.
The guys have split the traditional UI thread and the background thread down to the silicon.
Who knows, maybe the UI thread only needs access to the graphics card, and the background thread only access to network. It would save a lot of interrupt loops.
> Fans only need to run during intense use and the system can be quiet the rest of the time.
This is a bigger deal than it might seem. Fans running means that dust is accumulating in the machine at an accelerated rate, even if you have filters in place. The less the fans have to run, the less dust accumulates, and the longer cooling functions near-peak. A tower that sees sporadic heavy usage especially stand to benefit from fan minimization.
As has been mentioned elsewhere, devoting some cores to background tasks and reserving the others for higher-priority stuff also makes the system feel more responsive.
Long story short, segregating background work from foreground work means that Spotlight indexing won't cause the UI to become sluggish. I'm guessing it also tends to be more cache-friendly.
But separating the background work doesn't actually require efficiency cores. With 40 performance cores, you could still dedicate 4 or 8 of them to "background" processes only. As that article you linked shows, the background processes would simply complete more quickly without affecting interactive work on the machine.
performance cores are physically larger than the efficiency ones, by about 3-4x. I'm sure there are other tradeoffs cache size like too. So would you rather have 32 performance cores and 8 efficiency or 34 performance?
Consuming 1/10 as much power would also have thermal advantages for the die overall, and segregating them onto a separate L2 cache from the performance cores would help with that cache friendliness situation, different allocations of logic units means you can tailor them toward different kinds of workloads. . . I'm no CPU designer, so I really can't speak with authority here, but I can at least see several possible reasons why it would make sense even when you aren't worried about battery life.
In upper end extreme case, using 8 HE Core instead of 8 HP Core saves you 24 to 32W. Those 32W could allow more GPU usage or some other part of system such as ISP or NPU.
At the end of the day you will still need HE Core, they just aren't as important comparing to a Battery devices.
Not to mention Marketing, Die Size cost consideration.
The design challenge of high-performance CPUs today is heat. Too much heat induces throttling. Putting low-priority stuff onto efficiency cores reserves the power budget of the SoC for high-priority tasks.
I wonder what % of high end users actually work on battery only. For me it's mostly a desktop that I move between home and office. And like maybe 2% of time batteries.
This is what I would have expected - I suggested when the M1 was first mentioned that it might make more sense for the Mac Pro to just have a side wall of separate independent SOC stacks with independent on-thread memory that GCD could manage behind the scenes. This seems like a much more straight-forward way to throw more performance into the higher-end machines without making long-tail disruptions to the architecture.
The efficiency cores are essentially for when the system is idling or for the 95% case (browsing the web, watching YouTube and checking e-mails). If a task is performance-bottlenecked then it makes sense to go straight to maximum performance rather than any in-between combination as you're essentially either waiting for the computer or not (i.e. it probably doesn't make much sense to try to strike an efficiency balance for tasks that are stressing a minimal set of low-power cores when they need to be completed either way - and in a machine where there's no battery life to worry about).
Yes, and it's also because a 2 KB text today comes inside of megabytes of code that has to be executed, and some code is additionally very expensive to execute, as it's doing shady stuff in trying to fulfill all the promises of all ad companies involved, sometimes really hundreds for a single page.
There was a moment in web history when the same happened due to the flash-based ads which easily spent huge amounts of CPU for... nothing meaningful to the reader. Now the flash is gone, but the pattern repeated.
And even not counting all the ads, today's "hot" web development conventions are often making the CPU even hotter.
It's like there's this assumption that the web's simple serving of HTML was a compromise of the time rather than it's killer feature.
I don't want to sift through hundreds of images and outbound links from the ads you're trying to trick me into clicking, I just want the information I came for or I'm gone forever.
Generally speaking, Apple's systems throw work like web browsing to the performance cores. Browsing the web is actually super intensive for very short periods of time, increasing performance of rendering makes the system feel snappier. They mastered this art in the iPhone as they had to learn how to make snappy feeling devices that also sipped power, the easiest way to do this is to be done with processing as fast as possible so you can get back to idling on just the E-cores.
The E-core magic trick is that they are always running and also exist to prevent thread stoppers, tasks that force you to pause high-priority work such as network requests, push notifications, or memory management. This enables a machine where the performance cores can pound away at tasks uninterrupted and your cache can stay dedicated to the task at hand, preventing situations where you may have to dump branches mid-process, increasing performance consistency. Mathematically, it works out to where the P cores are ~5.2W each and then the E-cores are less than that combined, allowing the system to idle on just e-Cores and spin up and dedicate P-Cores to the UI and UX intensive tasks, keeping the thermals low and performance high. In a situation like the Mac Pro, I would expect much, much hotter P-Cores too as the thermals will allow.
A nitpick but arguably an important one for your point - this article argues that the efficiency cores aren’t for “browsing the web / watching YouTube / checking emails” but for non-interactive tasks like Spotlight indexing (based on GCD scheduling priority).
A performance-focused workstation would have more performance cores because it’s being used for more CPU-intensive work. If you don’t actually need that many performance cores, you’d buy a Mac Mini instead.
That seems a bit weird to me. Many programs are hard to multithread but I'd expect that if any given program can make effective use of 8 threads it can probably keep scaling to 16 and 32. The hard part is getting from 1 to 2 and 2 to 4. So I'd expect that in most cases having more than 8 high performance cores doesn't make that much sense.
A Mac Pro is likely to be dedicated to those workloads which ARE easy to thread - video editing and processing - or simply have multiple applications running simultaneously.
Someone who doesn't need more performance cores is unlikely to invest in a Mac Pro.
Macs are primarily well designed consumer devices, status symbols and objects of brand loyalty. These factors will push sales beyond a basic functional need.
I'm confused at this response. My thought is that instead of 32 high performance cores and 8 efficiency cores it might be better to have 8 high performance cores and ~100 efficiency ones which ought to be about the same area and power but provide a lot more throughput if software can utilize that level of parallelism.
There's a lot of software that won't scale to 108 cores, of course, but again I expect the amount of software that could usefully scale to 40 parallel threads but can't scale to 108 to be fairly small.
The performance cores seem to take maybe twice as much space as the efficiency cores, so I'm not sure the tradeoff would be worth it - the efficiency cores would need to be more than half as fast taking into account the parallelism to make the tradeoff worth it.
In a world where power and heat budgets don't matter at all you'd not have any efficiency cores, after all.
The general rule of thumb for reasonably sized cores, which both Apple's performance and efficiency cores are, is that performance tends to scale with the square of the transistor and energy budget due to N^2 structures like the bypass network, ROB, etc being the long poles. Now, for any given transistor budget there's going to be a non-linear tradeoff between performance and power governed by supply voltage versus process threshold voltage. And if you're testing a chip and using only one or the other of the two sorts of cores you're paying for the uncore and memory subsystem regardless of which cores you're testing which complicates things. But it would be really weird if power consumption wasn't super linear with performance, that's why "efficiency cores" have "efficiency" in their name.
Bad assumption. It's pretty difficult to write a multithreaded program that scales up to 32 cores. You have to carefully avoid all sources of process-wide contention. The only cases where it is easy to just keep adding cores are where the units of independent work are overwhelmingly large compared to everything else, like video compression.
No, it's certainly hard to write a multithreaded program that scales up to 32 cores. But it's also hard to write a multithreaded program that scales up to 8 cores. I'm saying that if we exclude the multithreaded programs that don't scale to 8 cores, most of them can be made to scale to 32 cores with reasonable effort.
I don’t really agree. You can get cpus where eight cores share an l2 cache and you can pretty much obliviously program those and get totally reasonable scaling because the contention penalty is the L2 access latency. There are no 32-core parts with that property, so you’re paying for at best L3 access latency and maybe main memory, possibly even NUMA node-crossing latency.
Are there CPU architectures where 8 cores share a L2 cache? I'm not aware of any though I can think of several where 4 cores share a L2 cache, though those tend to be pretty lightweight cores. More often you see issues with the L3 cache being shared between 8 cores or no as with Zen 2 versus Zen 3. And since only lightweight cores tend to share L2s this isn't really an issue in terms of replacing heavyweight cores with lightweight ones. Mostly, though when I think about sharing data between 8 cores I tend to be concentrating on making sure it ends up as S in MESI, MOESI, MESIF or whatever cache coherency protocol the chip is using.
Sure, POWER9 offers pairs of SMT8 cores sharing L2 caches, that's 16 thread contexts presented to software that are all sharing an L2 cache domain. Intel "Tremont" Atom has a shared L2 cache for a module of up to 4 cores. Maybe that is the one you allude to as lightweight.
Yes, there are clearly architectures where 8 threads share a L2$. L1$ too in both recent POWER SMT8 cores and SPARC Niagras. Yes, Intel Atoms were one of the examples I was thinking of regarding 4 cores sharing an L2$. One can quibble about what counts as a "core" when you're talking about POWER9 or Bulldozers but even being generous I don't think there's ever been a design where 8 cores all share the same L2$.
Even then, there’s the contention on memory access. If one of those easily paralleled tasks uses x% of your memory bandwidth, having more than 100/x CPUs won’t help.
Will TSMC 3nm be the second die shrink in a row where Apple does the unconventional thing and takes a reduction in power/heat instead of a performance increase?
>Compared to it’s N5 node, N3 promises to improve performance by 10-15% at the same power levels, or reduce power by 25-30% at the same transistor speeds.
Considering M1 only uses 20W max, you could add a lot of cores without any efficency savings and still be within the current Xeon 200W TDP envolope. Indeed 8 M1s to make 32perf/32efficency would only be 160W. I'm aware this is very rough, but I also imagine even 8 M1s are a fraction of the cost of one Intel Xeon W chip, so much more margin for Apple. Not good times for Intel.
AMD's high-end cores have a quarter-gig of cache. That uses a lot of power no matter how you slice it. AMD's Infinity Fabric interconnect uses 30-50% of total chip power. Apple will have similar communications overhead too.
Adding more memory lanes increases power consumption as does adding more PCIe lanes and other IO.
It definitely remains to be seen if their chips are significantly faster when scaled up. In any case, they wouldn't have to beat AMD or Intel to count it as a success. Simply not being terrible would be enough.
> AMD's Infinity Fabric interconnect uses 30-50% of total chip power.
It would use less if it was on the same process node as the rest of stuff though. They were fulfilling some obligation to Global Foundries by having them separately manufacture it.
>still be within the current Xeon 200W TDP envolope
Apple needs very high performance core designs that also work in the thermal envelope of a handheld phone or a fanless laptop.
As noted when the A14 launched, taking the power/heat cut at the 7nm node change resulted in better sustained performance for those chips
>The one explanation and theory I have is that Apple might have finally pulled back on their excessive peak power draw at the maximum performance states of the CPUs and GPUs, and thus peak performance wouldn’t have seen such a large jump this generation, but favour more sustainable thermal figures.
And I would assume that the GPU, which needs to be grown considerably too, might be on a second package, so the 8 M1s without GPUs would use less than 100W.
Could an informed reader briefly explain whether the current M1 chips are overly hard on their hard drives? I'm debating between 8GB and 16GB for an M1 Mac Mini. Between the two sizes my main concern is more about whether the swapping on the 8GB will kill the hard drive, rather than whether 8GB will be enough
The Mac Pro Mx CPUs illustrate the promise of ARM and Apple Silicon. When the Firestorm core is one of the fastest available but uses less than 5 watts, Apple can use a ton of them at once without massive heat and power usage.
The top end Mac Pro with 32 Firestorm on one chip is probably will have a GeekBench Multicore score approaching 60,000, triple the fastest Xeon based Mac Pro.
And it probably uses half the 300 watts of the existing Mac Pro, with correspondingly less heat. Hopefully Apple doesn’t use this as sn opportunity to go back to the whacky packaging of the Trash Can Mac Pro.
I don’t know about AIDA64, but Geekbench tests a variety of more real world or at least more varied scenarios compared to cinebench which just does CPU rendering.
Taking for example encryption, isn't it an unfair workload? Encryption is hardware accelerated nowadays, which means you won't get an apples-to-apples (heh) comparison of processors and computers.
Yes. I’m not sure exactly how geekbench implements it but that’s certainly a big downside. Perhaps geekbench runs it without any hardware acceleration (to just test a different type of workload) but it might just use whatever hardware available.
As an end-user I’m okay with that, because at the end of the day if things run faster I don’t really care how, especially for tasks that are ubiquitous.
As someone who planned to buy the new MacPro since the first rumors and only skipped it when I learned how much the prices had risen compared to its predecessor - and the entry level 8core Xeon wasn't that impressive either - I am so ready to finally upgrade my aged iMac. The 4+4 core M1 already gets glowing reviews, it will be very impressive with 16+4 or even 32+8 cores. I like those rumors very much!
I bought a 10-core iMac Pro, and it's a screamer - super-fast and super-reliable even after 16+ hours of use a day. So I don't regret that choice, even if it was the most expensive computer I've ever bought.
A 32+8 core M1 for a less insane price sounds like an easy upgrade choice. The lingering question is memory. 16GB won't be enough.
Really would like to know more about the GPU and Ram. At 128 gpu cores vs the 8 in the M1 you talking 16 times the potential GPU power which could put it up there near the top of discrete GPU's.
It takes a lot to scale a GPU up that far. Memory throughput requirements (the actually expensive part) increase linearly with number of cores. It will either cost a large fortune and use a Lot of power or be memory throughput starved (and cost a small fortune).
Tile based deferred renders help with memory bandwidth requirement (partly why UMA mobile GPUs and the M1 use it). The M1 has 16mb of cache coherent with CPU and GPU and then there is tile memory in each core.
That would be interesting for the Mac Pro since the kind of applications most people who would buy one would be running have probably never even been run on a tile based GPU, let alone optimized for it.
Curious about this too. Anything more than 16 GPU cores seems like it'd too big to fit on die (share with CPU cores and everything else), so may need to be a discrete solution.
Would Apple then make discrete GPU solutions using PCIe5? Makes sense for the Mac Pro given the edge cases for requiring multi-GPU solutions.
>Anything more than 16 GPU cores seems like it'd too big to fit on die
32 HP Core with 8 HE Core, 128 GPU Core give you ~650mm inclusive of L2 and SLC. Adding in I/O, NPU, ISP, VPU etc this chip should still be within the reticle limit of ~850mm2.
The A64FX has much smaller cores though using the similar design to the N1. Apple's HP Core are nearly double the die size ( Hence why it is faster, it is a trade off between transistor budget )
They have been talking up the advantages of unified memory so my guess is they will stick with integrated, but I would imagine they will end up supporting AMD or Nvidia discrete to some degree.
Thought this was the most important piece of info in that article. Could be a game changer for a number of applications considering the softwares and dev support Apple can provide.
My current Late 2019 MacBook Pro heats up like crazy and goes into thermal throttling whenever I'm doing wfh meeting with lots of video participants for a couple of hours. I really hope they dedicate some hardware for this type of use-case with plenty of sustained thermal headroom.
The latests Intel-based Macbooks aren't well designed around thermals, and if the software you are using to have this meetings is Microsoft Teams, which at least in Windows eats a lot of memory and CPU resources (and I expect the same on macOS), it's even worse.
I haven't tried yet, but AFAIK M1 Macbooks have a lot lower thermal issues, so that isn't a problem right now.
I have this same issue. Fans spin constantly throughout the whole meeting. Hot to the touch. I never realized just how much I hate laptop fans until WFH started.
> Codenamed Jade 2C-Die and Jade 4C-Die, a redesigned Mac Pro is planned to come in 20 or 40 computing core variations, made up of 16 high-performance or 32 high-performance cores and four or eight high-efficiency cores. The chips would also include either 64 core or 128 core options for graphics.
> For the new MacBook Pros, Apple is planning two different chips, codenamed Jade C-Chop and Jade C-Die: both include eight high-performance cores and two energy-efficient cores for a total of 10, but will be offered in either 16 or 32 graphics core variations.
Sounds like the Mac Pro processor would use 2x or 4x of the Jade C-Die, perhaps in a single package. This makes good economic sense because it may be challenging to justify a unique chip design for only a few thousand units / year.
Is this going to be inside an affordable box ie do I have to spend close to 5 figures for it? This is about as exciting as new cellular data speed announcements in a world where most data plans are extremely limited and quickly throttled one way or another. I still use Macs, but my resentment is building up from Apple's past choices e.g. thinness overrides every other feature, every component being solder on, ...
I would be very surprised if those Macs were offered at a much lower price point than the current top of the range.
They are selling the M1 MacBooks and Mini at a similar price as previous models with comparable number of ports, amount of ram, ssd. Only the CPU performance gain was for free ;)
The original title was "Apple readies 40 core M1". I was speaking of a desktop because I highly doubt that this is going into a laptop, Mac mini, or the non-pro iMac. If it is then I am happily wrong.
350 comments
[ 4.1 ms ] story [ 248 ms ] threadIt will be interesting to see how close the rumors are to fact when real hardware announcements happen. Given how fast the M1 in the MacBook Air is performing for tasks like video editing I wonder if they will really need to push to such extremes as putting 40 cores in a Mac Pro -- unless they plan to edit four 8K streams at once?
Or you could run up to 5 Electron apps simultaneously!
To give some perspective: the current version can go up to 28 Xeon W cores, 1.5TB or RAM, Radeon Pro Vega II Duo with 2 x 32 GB, ...
(~$54,000 USD with wheels)
"Titan S375 - Dual AMD EPYC Rome 7002 Series - Scientific Research Server PC up to 128 cores"
https://www.titancomputers.com/Titan-S375-Dual-AMD-EPYC-Rome...
The 40 core model is probably the replacement of the 28 core Xeon which is a $7,000 option.
https://data.mactechnews.de/493404.jpg
https://i2.cdn.turner.com/money/dam/assets/150514094302-appl...
https://imgv2-2-f.scribdassets.com/img/document/94353573/ori...
https://b2c-images.idg.zone/3292182_800x400_r.jpg
I still deal with 1080p and get annoyed at my hardware's limitations. If I could justify a Mac Pro and Apple's accelerator card, I'd get them in a heartbeat.
For certain applications, no matter how good the hardware you have is, there's always a desire for more, faster hardware.
Is that really a common scenario? I can see editors at the big studios using monsters like this (because the studios can afford it) but are we talking about a market of hundreds or thousands? Is this something a free-lance video person would use?
M1 is fast at editing, but not particularly fast when you factor in processing.
It's a bit mind-bending for those of us who remember the SGI Octane and Onyx days to think that 100x as much computation is available at 100x less power-consumption. (That's 10,000 times more efficient, for those following along at home.)
Between that and the almost infinite instantaneous computation power available in the cloud at the press of a few buttons (and sufficient capacity on one's credit card), we ain't in Kansas any more.
My theory is Apple will release the new Mac Pro with an updated Apple Silicon Afterburner card focused on pro video.
They will debut a new AI / ML compute "accelerator" focused on the data science space also under the Afterburner moniker.
https://news.ycombinator.com/item?id=26956886
Say what now?
The problematic part of that quote is where it claims that iPhones and iPads used to have Intel processors.
Interestingly though, Intel did at once point make very nice mobile CPUs (notably the XScale PXA270, which clocked in at a little over 500 MHz and had a halfway decent PowerVR graphics accelerator).
However I still wish they'd be able to get the bezel down to the size of an XPS 13
Right now I'm looking at a 2019 16" and 2018 15" models, they're pretty much the same size, the 16" model is just a couple mm wider. Probably the same applies for 13"->14".
For Windows support, I thought Apple said something along the lines of: "If Windows wants to add support, we are fine with it" - but I could be mistaken.
This new design will likely replace that one in Apple's lineup, but it never went away.
Like I said before, I skipped the M1 and its seems to be paying off.
[0] https://www.youtube.com/watch?v=iiCBYAP_Sgg
Seriously, WTF? Where is it?
1: https://i0.wp.com/lemmymorgan.com/wp-content/uploads/2020/01...
The heat pipe is the part just above and to the left of the fan.
(Video mentions 2019 model, which is a minor revision based on this)
The downside is it's harder to get decent airflow over the fins. The casing is probably designed to direct the airflow, but I'd have to see all parts together to figure out how this works.
EDIT: the lower case appears to have some ducting on it, see this example: https://www.ebay.co.uk/itm/324259489798
Provides for heat dissipation over broader surfaces. This is normal.
And maybe it is a mistake. That happened before and they iterated over it and turned the design into an advantage most of the times.
Judging and say "does nothing" is very naive in my opinion and I am pretty confident that everything in that designed is planned pretty accurately.
Apple is also known for not admitting its design failures.
1: https://www.zdnet.com/a/hub/i/2020/11/20/ab77bd9b-edc0-419f-...
The idea that a cooling fan needs to “blow on” the thing that it’s cooling is just dumb. As long as the path from the fan to the CPU is sealed, it makes no difference at all how far away the fan is. (Think of a wing being tested in a wind tunnel. The fan that’s making the ‘wind’ doesn’t need to be placed right up against the wing for this to work!)
A heat pipe gives you more layout flexibility because it lets you place the fan somewhere where it can’t generate airflow directly over the CPU. Given that a heat pipe can’t be 100% efficient at transferring heat, a design with a heat pipe is actually less efficient, all else being equal.
> I am pretty confident that everything in that designed is planned pretty accurately.
Apple was designing stuff that considered the airflow THROUGH THE CASE back with the first Macintosh, the 128K. You could even get a silly looking chimney that used convection to pull even more air through the case with no other modifications.
All Macs with air cooling, not just the laptops, are designed with internal airflow in mind and always have been. You do not have a little fan blowing directly on the CPU (from where? to where?) and expect it to do anything.
I don't know who this fellow is, but next he can take the top off his cylinder head and then be upset the engine doesn't work. From first appearances he is an absolute bozo.
He's pretty much the main figurehead of the right to repair movement from the consumer electronics side of things. He has run his own independent repair shop doing board level repairs for years in NYC.
https://www.fighttorepair.org/ is him.
The person who bought it.
There’s no particular reason the fan should be physically next to the CPU or heat sink. Its job is to draw air into the computer, across the heat sink, and then back out.
You generally don’t want the air intake and outflow to be too close to each other since that would typically cause heated air to be drawn back in, making the cooling a lot less effective. And you mostly want the air flowing in a channel (rather than blown all around inside), so you can expel most of the heated air.
So, e.g., you have intake on one side of the computer, a channel for the air to flow inside the computer, and a fan at the other side of the channel, pulling air through.
This will be effective if the heat sink on the CPU is anywhere along the channel. There’s no reason it needs to be next to the fan.
A piece of copper, BTW, is a way to move heat from some place that doesn’t have good exposure to the air channel to a place where it does. You only need to do that if the heat sink doesn’t already have good exposure.
To do this properly, I think you’d want to analyze the design and then run some tests to see under what circumstances heat builds up.
Lost a little respect for that guy. He’s a smart guy, so I know he understands this stuff. You’ve got to wonder why he’s bullshitting his audience like this. Seems clear to me this video is about promoting himself rather than discovering or explaining flaws in the cooling system of Apple laptops. (Or maybe I’m giving him too much credit, and he really doesn’t understand this stuff. Either way, it’s not a good look for him.)
Even if they continue to force the anti-feature that is the Touch Bar onto users, I'd immediately buy an M1 MBP with MagSafe and actual ports.
Yes, I did use the SD card reader from time to time, but most serious cameras use other cards (my older camera uses compact flash, and the newer one uses XQD). It makes no sense for Apple to integrate readers for those, and my XQD/CF readers also read SD cards, so the card reader is useless.
All of my mobile peripherals at this point are USB-C, so I don't care about USB-A ports either. And when I'm at home, I use a dock, which has all of the ports I need.
Honestly I'd much rather that they keep the laptops thin instead of adding ports I'm not going to use.
The vast majority of people these days just use their phones for cameras, and don't use point-and-shoots or other things with full-size SD cards.
So really SD cards cater to just the mid-range camera market, where people use mirrorless/DSLRs but not high-end versions that often use faster cards.
I don't think it's a good idea for Apple to make everyone's laptops bulkier so a small minority of people can avoid using a $17 adapter like this:
https://www.amazon.com/gp/product/B07NW8RPYN/ref=ppx_yo_dt_b...
With that said, I have cameras that use SD cards and don't see the need for an SD port. I have usbc adapters that are not much bigger than the SD card that I can use on my MBA or iPad. With LR, I can basically load my days pics on either device and have them available everywhere.
I'm very skeptical they are adding back SD, so we'll see. I think MagSafe is going to be something like what came to the latest iPhone and not what people think of traditional MagSafe. Or maybe they'll add something to the usbc port that works like MagSafe if using an Apple cable.
It's finally, years later, getting merely not infuriating to be stuck with only USB-C ports.
Single port for display + charging is very nice, when it works.
That iPhones and most iPads still don't use it is annoying. Having two kinds of charging cable around, years after Apple went all-in on USB-C for laptops, sucks.
[EDIT] two kinds of charging cable and still needing to buy & carry two kinds of adapters/dongles for everything. Ugh.
The most recent non-pro iPad does have USB-C and I expect whenever they get around to redesigning the lowest cost iPads they’ll also move to USB-C.
Man, my 2017 MacBook Pro's USB-C charging cable (the official one, not third party, and in pristine condition) constantly disconnects itself to the point where I'm basically traumatized from the annoyance of it now. It only provides power if it's ALL the way in, and the damned thing just pushes itself out half a millimeter and cuts power repeatedly.
I really want to like the generic nature of USB-C charging, but proper magsafe with a strong magnet is just miles better (third party magnet adapters for USB-C both break the spec and also tend to have very weak magnets which can cause damage from sparking at the port).
Yeah, well, it's not.
> Yes, but how about the _inside_ of the USB-C port?
It's pristine, like I said elsewhere already.
Yesterday I moved it around. And noticed what you described. The plug will come back out by a millimeter. And then it won't charge.
Maybe just a bad design of the usb-c port then.
I don't have anything helpful to offer but I feel your pain. Reading your comment almost brings on PTSD from remembering what it was like.
I actually restored my 2015 Macbook Pro and have been so happy and content using that in the past two years or so. Not sure about the other sizes, but my 13" from 2015 does everything better than the one from 2017. Performance, keyboard, connectors. None of the overheating bullshit where the fan is constantly going and performance is throttled.
I felt the same until my cat pulled my laptop off my desk by the cord. It landed on the corner with the USB plugged in, destroying the port on the laptop itself. Probably knocked a few hundred off the resale value. The little shit.
They never quite got their Intel laptops there, though. (Especially running MS Teams, Slack, and Chrome!)
It’s kind of funny that, now that they can make a laptop that really does last all day on one charge (according to tests of M1), they’re switching back to MagSafe.
There's this stupid dream that the computer needs to be light and fanless and thin and sexy. It doesn't need to do that. I'm happy if it doesn't melt while trying to run some software I need to do my job.
Modern computing is a sad place :(
lol, it's not only browsing though but web-infested Electron apps as well. Used draw.io (which is otherwise a fine diagramming app) the other day, with fans spinning up all the time. It's sad to see that draw.io was (and internally largely remains) originally a Java (Swing) app, seemingly only ported to a fscking browser frontend as a fashion statement when the exact same things it does I could do almost 20 years ago on a PowerBook using OmniGraffle.
Man is that an understatement. My work MBP - a 16" i9 - may as well be a desktop.
Whats up with photo/video people gatekeeping “serious” vs functional
Your point might have some merit but SD will hold out for the next 5-8 years that purchasers of this machine will expect
That said, I'm glad to see Apple has finally realized some of the ways they've been alienating their customer base, and they appear to be walking back some of their most unpopular changes.
I’m looking forward to having that in the new processor and with touchbar
They are acknowledging the folly of their ways as they have gotten rid of the butterfly keyboard, adding magsafe back, and adding sd cards back
They are making the least contentious and perfect machine
My "perfect machine" could also play DVDs, but that may be getting a bit contentious.
Regardless, the point remains that high-end cameras often use high-end memory cards that are not SD cards.
Have you heard of these obscure websites called Amazon or eBay.
They have tens of thousands of cables in every configuration you need.
Last time I needed a cable on short notice before a gig I had to settle for an adapter, and was really lucky that it worked. I've had other seemingly simple pin-to-pin adapters not work on me before.
Next time I'm in a jam I'll run down to my local ebay though, thank you for this super helpful reply.
From these reports, the next revision to the Air is going to get a bit of a redesign, colors, faster CPU, etc., but "Air" is still almost certainly going to mean the line that prioritizes being thin and light. You will still have a Mac laptop that gives you what you want. But people who do want more ports will, for the first time since Mac laptops went USB-C, also get what they want. (Or at least get closer to it.)
That's the only change that my next laptop isn't a ThinkPad
https://9to5mac.com/2021/04/21/apple-ransomware-leak-new-mac...
looks like it has physical keys.
I think that's the right side of the laptop, so from the bottom up, it is shift, return, \, delete, and the connectors for the fingerprint sensor.
hallelujah
The fact the OS runs a dozen network services without being able to deactivate them is (the final straw) the pushed me back into the Linux camp.
I want to both beef up the memory/processing power of my laptop and switch to Linux (which comes with built-in performance bonus of native Docker). So I am headed to ThinkPad anyway, but there's a chance Apple could keep me locked in if they make a nicely-powered proper dev machine with no damn touch bar.
Migrating to Big Sur has been a pain in the ass even though I am still on an Intel model. Tons of Python language versions and packages are a total PITA to get built on Big Sur.
To be fair this seems to be a pretty Python-specific issue, but the inability to do much about it really has dug the feeling into me that I don't own/control my machine, Apple does.
The touch bar is handled by the T1/T2/M1 chips in the laptops. It could be that just cramming a T2+OLED into a keyboard may have made it way too expensive for a keyboard. Plus does Bluetooth even have enough bandwidth to handle sending screen data that size reliably?
It is a niche item and it does suck but for some reason they keep tacking it on...
That's because changing the chassis is too expensive, but when they redesign it, you'll see it most likely without a touchbar.
I'm not a TB hater (more indifference), and I still think it's going away :)
Looks like you won't have too long to wait:
Apple ransomware leak corroborates 2021 MacBook Pro ports: HDMI, MagSafe, SD card slot https://9to5mac.com/2021/04/21/apple-ransomware-leak-new-mac...
Not exactly sure if I'll end up buying a MBP later though. I basically only use my Mac for recording audio, and once you get your mic set up , your not too portable anyway
When they first announced the transition last year, it was implied that some products will still get refreshed during the 2-year transition period. But based on this article and common sense, there is minimal incentive to do an Intel refresh. Refreshing both discourages buying of Apple silicon-powered Macs and lessens the performance jump that Apple is so eager to publicize.
For example, Tiger Lake H-series was just announced/released and it'd logically target the Macbook Pro 16.
Yes, Tiger Lake H-series was just announced, but it's likely Intel proposed selling this chip to Apple 2.5 years ago, and Apple didn't like the roadmap, and said No Thanks.
That's how much time supply chain changes like the move to apple silicon chip will need - I wouldn't be surprised if Apple already knew of all of Intel's roadmaps 5 years into the future, and decided, nope. No thanks. Not giving any more money to Intel.
From this point of view, you can tell, it's detrimental to Apple to go back to Intel now, if they wanted to use the Tiger Lake H they would have used the threat of M1 as a bargaining chip to lower the pricing on upcoming chips. It's obvious they didn't do that.
For what it's worth, once I got the physical escape key back, I've come to really like the Touch Bar.
It might not be a 19" rackmount chassis like the Xserves of yore, but it's hard to deny Apple hardware will continue to find its way into real datacenters.
The high end Mac mini SKU mentioned in TFA sounds great for this application, especially now that they have 10Gig Ethernet.
ARM might replace Intel, but it won't be branded Apple, that's for sure.
I'd be a bit concerned to write something with parallel processing, only to have an overheating issue denied by Apple for years. Butterfly keyboard flashbacks...
I don't dispute that some applications are hard to parallelize. However, in my day to day work, I've found that parallel operations are often not that complicated to introduce.
The bigger issue is that CPUs are now so fast I find myself blocked by IO far more often than I'm blocked by actual processing.
> From production rendering to playing hundreds of virtual instruments to simulating an iOS app on multiple devices at once, it’s exceedingly capable.
> Increased speed for Logic Pro X, MATLAB, Adobe Photoshop 2020, Autodesk Maya, Wolfram Mathematica Build Time, Final Cut Pro X
> Tasks like animating 3D film assets, compositing 8K scenes and building lifelike gaming environments.
This is the target market - and pretty much everything listed here tends to be multi-threaded. Making multi threaded applications aren't easy, but for high performance tasks all these use-cases have to be multi-threaded where possible or they are just too slow.
I think that people grossly underestimate the amount of processing that can be done on video. Reasonably sized filters, something like deinterlacing, high quality transforms. Swamping a modern CPU in terms of memory access or processing isn't too hard.
Yes, if you just use AOM directly you'll have a hard time saturating the core with work.
However, if you split your video up into scenes and start an AOM instance per scene, it becomes trivial to outperform the 12 core ryzen with a 128 core machine. The main bottleneck becomes memory bandwidth.
This is what the av1an project is doing https://github.com/master-of-zen/Av1an
All that said, you need videos long enough for there to be a benefit here. The Ryzen will still win if you are talking about making 10 second av1 gifs.
Mac Pros will have either 16 performance/4 efficiency, or 32 performance/8 efficiency.
Jobs specifically hated that, and it was indeed the Mac II where Gassee and others pressured him into producing a Mac with more expandability.
I'd say if you purchased a Mac Pro, it's still the same. If you find yourself not having enough power, you bought the wrong machine to begin with (probably modern threadripper pro would be what you should have been looking for)
If the computer satisfies all your performance needs, what use is it to "feel frustruated" over something that was expected to come?
Don't feel Fear of Missing Out. The Mac Pro didn't change. Your feelings about it changed because of marketing and advertising (ie apple trying to persuade you that you need the M1, when in fact your computing needs are being taken care of already)
Wait for the next generation after that current M1/M1x mac pros and go for the next generations when apple comes out with 128 Cores and 256G of ram.
One constant of computer industry, there will always be the next chip, the next computer, and by then, the fancy M1 Mac Pro will become obsolete. Exactly like how nobody wants a 486 computer anymore but when it was released it was the hottest thing on the planet. Its day will come when everyone says "oh that M1? that was a slow POS that is now ewaste"
1) The details were all speculation. 2) There was no other option if you needed more horsepower. 3)_I'm happy with the current Mac Pro, but its going to be phased out of software support before its time, just like the Motorolla.
So it's more about the software support being phased out than new hardware coming out.
If you want the software support, wait until the support dies/runs out, and then buy the newest computer at that time, when you need it.
You'll find that apple will drop support for the M1 too in the future. Its day will come when Apple decides the neural engine is too outdated or the architecture is too old and will stop support it as well. That's a fact, not speculation.
This is precisely why I as a consumer do not buy Apple products. Apple has a long history of dropping product support with no upgradeable path forward, only buying new hardware.
This is also, in my speculation, the reason Apple is one of the most valuable companies on the planet, even though they are primarily a consumer products company. There's a constant internal urge for the newest shiniest product (which consumers of all products feel), but also the very real external requirement to buy new if you want to work with anything new.
My best anecdotal experience for this was when my uncle contacted me because his wife bought him an iPod Touch, but his Mac was telling him he needed a newer version of iTunes (I believe he bought the Mac 4 years prior). It turned out however that the new version of iTunes couldn't be installed on his "outdated" Mac because it was bundled with an OS update that his machine was too old to receive. I meanwhile was able to load iTunes under Windows on a machine twice the age of his Mac and connect the Touch without issue. He didn't want a new Mac, didn't need a new Mac, but chose to buy a new one to work with the rest of his ecosystem.
It's not speculation to know that whatever computer is bought this year will be bested by what comes next year.
Comments like this are always just so empty to me. This person was looking for a more powerful Mac, so suggesting they should have been looking for a CPU that's never going to be available for use in a Mac is just grandstanding. This person had been waiting for years for a product from Apple to replace the worthless trashcan systems. How your suggesting of something never to be available is helpful eludes me.
My suggestion was if there wasn't enough power, and the person needed a computer (which is the only reason one would buy a mac exactly during this changeover point), then a threadripper pro running linux would be the way to go for many workloads.
The alternative is to just wait a few months and see what comes.
Except for all of the software that doesn't run on Linux.
>The alternative is to just wait a few months and see what comes.
If one did this with Apple, one would never buy a product.
This is the reality really, if you actually wanted power then the Mac Pro just wasn't what you should have bought. Threadripper + Nvidia cards would take you way further for far less and with a clear sustainable upgrade path.
Also really after the past 10 years of Apple's pro moves with their last 2 "workstation" machines being discontinued it seems very risking investing so much money in their pro line.
Everyone then bought the current Mac Pro, but were disappointed that nothing else really fits into the slots aside from the things that Apple offered themselves (Nvidia cards won't work in the current Gen Intel Mac Pros.)
What makes anyone think the next Apple Silicon based Mac Pro will be any different?
Wanting a more modular mac pro is more like wishful thinking or wanting something to be different when reality is clearly saying something.
It's a little upsetting to see even the 12-core version get smoked by the M1 Mini, but it's from 2013 so what do I really expect.
Still nice and quiet and compiles iOS apps just fine.
So far the answer seems to be, there is no getting around these CPU issues through tailoring the compiler flags etc. and the best people can get is amelioration of the problem, not getting it back on par with normally inferior Intel chips. Again, the issue is that the Threadripper stuff is eating unreasonable CPU and the Intel stuff, normally not as good, is not doing that.
I suspect what you're looking at there is a CPU architecture that's getting optimized for specific conditions, which my plugins don't meet. There seems no fix other than 'you should be playing Mass Effect on this computer instead'.
I'm sure that it would destroy the Intel computer if you're doing that, but generalizing that to all conditions seems not justified.
Buy a mac, sell it after 1 year (many stores offer a buy back program for this exact purpose), get another new mac.
Easy. Done.
When the "new" Mac Pro was announced, the shift to ARM processors for MacOS was at most rumored and it also wasn't obvious that such a CPU would perform well.
If people knew this was "going to come", they would likely not have invested a significant amount of cash into a platform that was already on its way out.
> One constant of computer industry, there will always be the next chip, the next computer, and by then, the fancy M1 Mac Pro will become obsolete.
Unlike the incremental improvements you get with each new Intel CPU generation, this is quite a disruption. The PPC Macs soon were obsolete after the Intel Macs came about. All the efforts went into the new platform.
> Its day will come when everyone says "oh that M1? that was a slow POS that is now ewaste"
Things don't move that fast anymore. Take a 10-year-old Intel CPU, it's still quite capable. That certainly wasn't the case 10 years ago, or 20 years ago.
I kid, obviously. I've never seen anybody refer to the PowerPC as Motorola machines even though Motorola was part of the PowerPC alliance.
And third-parties will supply upgrades as the upgradable parts are roughly standard, so the Mac Pro is likely to be a completely usable machine for nearly a decade to come.
After that you’ll still get years on older MacOS, Linux, or Windows. Intel Macs can run both of the latter.
https://news.ycombinator.com/newsguidelines.html
Feels like wanting to be both on the cutting edge of performance, and running/developing for a Mac, is damn pricey. However, I'd rather have it this way, than sit with my current machine being the best you could get. In some ways I am gratified that Apple is busy making last year's news feel really inadequate.
You can be damn sure I'm devoting a lot of effort to promptly supporting the new architecture. I've already got literally hundreds of M1 builds for my Audio Unit plugins out. I chose to produce native versions of EVERY product, no matter how old it was: rather than make a new product and only release that for M1, I treated it as the same as the 32/64 bit transition, and re-did every product in the whole library.
If we're going to get absurdly capable Mac computers, I mean to be categorically included in that equation.
For now, I'll enjoy my very costly iMac Pro that's still very capable… AND the M1 Macbook Pro at a fifth of the price that outperforms the iMac Pro at a few things already.
- 20 total, 16 high performance, 4 efficiency
- 40 total, 32 high performance, 8 efficiency
Which is kind of interesting given that the 8 core M1 is split half/half at 4 high performance, 4 efficiency.
Sounds like perhaps MacOS doesn't typically have more than 4 cores worth of low-priority work, regardless of the system size/workload?
How many is optimum is an interesting question, I would assume the apple peeps have a good idea and are on the money with this design.
At some point it's more efficient to have a single performance core going than a number of efficiency ones.
They do up to 8 on the big CPU according to rumor, so there's obviously some room there.
There's also plenty of time where the system has low intensity use or is idle. The efficiency cores let things like receiving notifications or checking e-mail happen for very little power draw. Fans only need to run during intense use and the system can be quiet the rest of the time.
Who knows, maybe the UI thread only needs access to the graphics card, and the background thread only access to network. It would save a lot of interrupt loops.
This is a bigger deal than it might seem. Fans running means that dust is accumulating in the machine at an accelerated rate, even if you have filters in place. The less the fans have to run, the less dust accumulates, and the longer cooling functions near-peak. A tower that sees sporadic heavy usage especially stand to benefit from fan minimization.
See, for example, this article from yesterday: https://eclecticlight.co/2021/05/17/how-m1-macs-feel-faster-...
Long story short, segregating background work from foreground work means that Spotlight indexing won't cause the UI to become sluggish. I'm guessing it also tends to be more cache-friendly.
At the end of the day you will still need HE Core, they just aren't as important comparing to a Battery devices.
Not to mention Marketing, Die Size cost consideration.
The design challenge of high-performance CPUs today is heat. Too much heat induces throttling. Putting low-priority stuff onto efficiency cores reserves the power budget of the SoC for high-priority tasks.
https://arstechnica.com/gadgets/2021/05/apples-m1-is-a-fast-...
The efficiency cores are essentially for when the system is idling or for the 95% case (browsing the web, watching YouTube and checking e-mails). If a task is performance-bottlenecked then it makes sense to go straight to maximum performance rather than any in-between combination as you're essentially either waiting for the computer or not (i.e. it probably doesn't make much sense to try to strike an efficiency balance for tasks that are stressing a minimal set of low-power cores when they need to be completed either way - and in a machine where there's no battery life to worry about).
I know people say this all the time, but the reality is modern web browsers are incredibly expensive processors.
There was a moment in web history when the same happened due to the flash-based ads which easily spent huge amounts of CPU for... nothing meaningful to the reader. Now the flash is gone, but the pattern repeated.
And even not counting all the ads, today's "hot" web development conventions are often making the CPU even hotter.
It's like there's this assumption that the web's simple serving of HTML was a compromise of the time rather than it's killer feature.
I don't want to sift through hundreds of images and outbound links from the ads you're trying to trick me into clicking, I just want the information I came for or I'm gone forever.
The E-core magic trick is that they are always running and also exist to prevent thread stoppers, tasks that force you to pause high-priority work such as network requests, push notifications, or memory management. This enables a machine where the performance cores can pound away at tasks uninterrupted and your cache can stay dedicated to the task at hand, preventing situations where you may have to dump branches mid-process, increasing performance consistency. Mathematically, it works out to where the P cores are ~5.2W each and then the E-cores are less than that combined, allowing the system to idle on just e-Cores and spin up and dedicate P-Cores to the UI and UX intensive tasks, keeping the thermals low and performance high. In a situation like the Mac Pro, I would expect much, much hotter P-Cores too as the thermals will allow.
https://eclecticlight.co/2021/05/17/how-m1-macs-feel-faster-...
<personal rant>
I'd swear that Spotlight indexing would prefer a freaking supercomputer when it runs.
My whole system just starts to feel sluggish, especially if I just moved a bunch (say 1-5GB) of files onto my system and not it wants to index them.
</rant>
<math>S_\text{latency}(s) = \frac 1 {(1 - p) + \frac p s}</math>
Someone who doesn't need more performance cores is unlikely to invest in a Mac Pro.
There's a lot of software that won't scale to 108 cores, of course, but again I expect the amount of software that could usefully scale to 40 parallel threads but can't scale to 108 to be fairly small.
In a world where power and heat budgets don't matter at all you'd not have any efficiency cores, after all.
Bad assumption. It's pretty difficult to write a multithreaded program that scales up to 32 cores. You have to carefully avoid all sources of process-wide contention. The only cases where it is easy to just keep adding cores are where the units of independent work are overwhelmingly large compared to everything else, like video compression.
[0] https://en.wikipedia.org/wiki/MESI_protocol
The MBP M1x will have 8 Firestorm cores, 2 Icestorm cores and 16 or 32 GPU cores, support more ports, and up to 64Gb RAM.
These are going to fly, probably close to double multicore performance over M1, and up to 4X graphics performance.
>Compared to it’s N5 node, N3 promises to improve performance by 10-15% at the same power levels, or reduce power by 25-30% at the same transistor speeds.
https://www.anandtech.com/show/16024/tsmc-details-3nm-proces...
It would certainly help pack that many cores into a small space without throttling.
AMD's high-end cores have a quarter-gig of cache. That uses a lot of power no matter how you slice it. AMD's Infinity Fabric interconnect uses 30-50% of total chip power. Apple will have similar communications overhead too. Adding more memory lanes increases power consumption as does adding more PCIe lanes and other IO.
It definitely remains to be seen if their chips are significantly faster when scaled up. In any case, they wouldn't have to beat AMD or Intel to count it as a success. Simply not being terrible would be enough.
It would use less if it was on the same process node as the rest of stuff though. They were fulfilling some obligation to Global Foundries by having them separately manufacture it.
Apple needs very high performance core designs that also work in the thermal envelope of a handheld phone or a fanless laptop.
As noted when the A14 launched, taking the power/heat cut at the 7nm node change resulted in better sustained performance for those chips
>The one explanation and theory I have is that Apple might have finally pulled back on their excessive peak power draw at the maximum performance states of the CPUs and GPUs, and thus peak performance wouldn’t have seen such a large jump this generation, but favour more sustainable thermal figures.
https://www.anandtech.com/show/16088/apple-announces-5nm-a14...
The top end Mac Pro with 32 Firestorm on one chip is probably will have a GeekBench Multicore score approaching 60,000, triple the fastest Xeon based Mac Pro.
And it probably uses half the 300 watts of the existing Mac Pro, with correspondingly less heat. Hopefully Apple doesn’t use this as sn opportunity to go back to the whacky packaging of the Trash Can Mac Pro.
You can see the full description here (https://www.geekbench.com/doc/geekbench5-cpu-workloads.pdf) but there’s compression, encryption, compiling, and several other tests.
As an end-user I’m okay with that, because at the end of the day if things run faster I don’t really care how, especially for tasks that are ubiquitous.
A 32+8 core M1 for a less insane price sounds like an easy upgrade choice. The lingering question is memory. 16GB won't be enough.
What does the time have to do with it? It's not like a car with moving parts.
https://developer.apple.com/documentation/metal/gpu_features...
Would Apple then make discrete GPU solutions using PCIe5? Makes sense for the Mac Pro given the edge cases for requiring multi-GPU solutions.
32 HP Core with 8 HE Core, 128 GPU Core give you ~650mm inclusive of L2 and SLC. Adding in I/O, NPU, ISP, VPU etc this chip should still be within the reticle limit of ~850mm2.
Block diagram over a die shot: https://3s81si1s5ygj3mzby34dq6qf-wpengine.netdna-ssl.com/wp-...
I haven't tried yet, but AFAIK M1 Macbooks have a lot lower thermal issues, so that isn't a problem right now.
> Codenamed Jade 2C-Die and Jade 4C-Die, a redesigned Mac Pro is planned to come in 20 or 40 computing core variations, made up of 16 high-performance or 32 high-performance cores and four or eight high-efficiency cores. The chips would also include either 64 core or 128 core options for graphics.
> For the new MacBook Pros, Apple is planning two different chips, codenamed Jade C-Chop and Jade C-Die: both include eight high-performance cores and two energy-efficient cores for a total of 10, but will be offered in either 16 or 32 graphics core variations.
They are selling the M1 MacBooks and Mini at a similar price as previous models with comparable number of ports, amount of ram, ssd. Only the CPU performance gain was for free ;)