Bloomberg: Apple Working on Next-Gen Apple Silicon Chips for MacBook Pro, iMacs, and Mac Pro Due to Launch Next Year

[Boil]

32 cores...?

Is that 32 Performance cores & 4 Efficiency cores...?

Or might it be 28 Performance cores & 4 Efficiency cores, for a total of 32 cores (while giving us 28 Performance cores to replace the 28-core Xeon)...? ;^p

64 & 128 core options on the smaller Mac Pro, that is exciting...! Octane X should be ridiculous on these options...!

The question is what form this Mega SoC might take...?!? Monolithic SoC...? SiP (System in Package) with GPU cores as a different piece of silicon than the rest of the "traditional M1-style" SoC, but still connected by a large bank of HBM3 on the package, and possibly feeding a tiered memory, secondary DDR5 for larger needs...?

Like, 32 or 64GB (or even 128GB) of HBM3 on the package, and up to 1TB with eight DDR5 DIMM slots...?

 

[fourthtunz]

You were expecting them to perhaps stick with 8?

It’s always seemed pretty obvious to me that the core count would keep doubling up, as they scaled up to workstation level.

Here’s a prediction they didn’t mention, they won’t stop at 32 core CPUs.

No, of course not. Just glad to see this rumor.

 

[pankajdoharey]

Intel and AMD both are in a soup with their Thermal throttling so called CPU's , they both need to seriously think how to counter this, especially sad for AMD which just recently started to recover from a decade old slumber, will be again set back producing chips that are Industries second fastest and seriously hot. Intel has no hope anymore they cant get their 5 nm act right.

 

[TrueBlou]

No, of course not. Just glad to see this rumor.

It will be interesting, exciting, awesome, to see what a (dual/quad?) 32 core CPU and 128 core GPU can achieve when they do eventually roll along. I’m expecting my jaw to hurt from hitting the floor.....again

 

[MrDerby01]

Might be a dumb question.. Would Apple be able to have their Mx chip with a video card? Could do some serious gaming again.. I want to buy, but want more power.

 

[mjtomlin71]

 

[cgsnipinva]

Intel and AMD both are in a soup with their Thermal throttling so called CPU's , they both need to seriously think how to counter this, especially sad for AMD which just recently started to recover from a decade old slumber, will be again set back producing chips that are Industries second fastest and seriously hot. Intel has no hope anymore they cant get their 5 nm act right.

While this potential industry shift has surprised most of the legacy companies - do not count them out. AMD and Intel in particular have a lot of capital and manufacturing base. Not to mention they do have top flight engineering talent as well. The thermal issues they have now are more related to the x86 (CISC) architecture and legacy backwards compatibility that has to be carried into new products.

Apple's success in this area will mostly impact Apple near term. The real competition will come when Intel, AMD, and I believe NVIDIA start pushing towards the ARM market. Never discount a threatened animal with a lot of resources.

 

[fourthtunz]

It will be interesting, exciting, awesome, to see what a (dual/quad?) 32 core CPU and 128 core GPU can achieve when they do eventually roll along. I’m expecting my jaw to hurt from hitting the floor.....again

Heh yes, getting ready for mind blown. Again!

 

[patent10021]

My $4K Apple Store gift card is waiting 😍 I held out all 2020 for something like a M2 MBP 32GB.

 

[gk4]

well if you believe that the Mac is a superior machine to the other pcs out there then its not a tax.
Same as high end anything, better just costs more.

That's completely wrong. You can both believe the system is good and accept that Apple is applying a large markup to components like RAM and storage.

 

[leman]

There's no reason why the Mx chips for Mac Pro/27" iMac need to use LPDDR (since power is less of an issue) and AFAIK the only thing "special" about having on-package RAM is that the ultra-short memory busses can run a bit faster. You could still have "unified memory" with external RAM modules, but DDR4 would just have to run slightly slower... if only there was an even faster DDR5 coming real soon now. Which, of course, there is...

The more the cores, the more bandwidth you will need. I am not sure that regular slotted RAM will fit the bill here...
Multi-channel DDR5 sounds like an option (8 channels would give you 400GB/s), but I am not sure whether it's a feasible solution for an iMac, while for a Mac Pro that will most certainly be not enough. Stacked wide-interface RAM chips sound more realistic, but that of course means no user-upgradeability...

 

[laz232]

I am curious what tools you use about which you can say that with such confidence? I have found many tools that do not have macOS versions that do have iOS/iPadOS version (meaning they now run on Apple Silicon macOS systems), and many more that have moved to the web/Electron. If Apple has machines that are seriously competitive with the highest end systems, I would not be surprised to see other software ported either native to macOS or to Windows on ARM as a VM on macOS.

Actual real software (like the stuff used to design the Apple Macbook PCB):

This is only availabe for Windows - no chance of Windows for ARM versions:
Altium Designer (6500 euro / seat)
Tektronix SignalVu (needed to operate my 10 000 USD real-time spectrum analyser)
Zuken CadStar
Sonnet Suite EM (unless you buy the Gold version, which has RedHat Support - but not redHat on ARM)
Various other circuit simulators

I'm often at client sites - so a laptop is necessary - don't want to use 2 laptops and prefer Mac side of things for file management, admin workflow etc.

 

[cmaier]

Intel and AMD both are in a soup with their Thermal throttling so called CPU's , they both need to seriously think how to counter this, especially sad for AMD which just recently started to recover from a decade old slumber, will be again set back producing chips that are Industries second fastest and seriously hot. Intel has no hope anymore they cant get their 5 nm act right.

AMD has a culture of punishing the engineers for their success. “Opteron is pretty good, but it took you 24 months instead of 21 months to get it done, so it was a complete failure. You now report to some dude we just hired in Colorado , because he’s got some great ideas.”

 

[Unregistered 4U]

AMD and Intel in particular have a lot of capital and manufacturing base. Not to mention they do have top flight engineering talent as well. The thermal issues they have now are more related to the x86 (CISC) architecture and legacy backwards compatibility that has to be carried into new products.

AMD and Intel are still stuck to a huge base of users that ONLY want incremental changes to what already exists. I’m sure at both companies there’s a large backlog of awesome ideas “If we could only get the developers on board” or “If we could only get the OEM’s on board”. One or the other could make strides in an industry leading direction that breaks some backwards compatibility but offers staggering performance... and the other would just figure out a way to provide a more meager way to improve performance BUT while maintaining backwards compatibility. You can guess which direction the industry would lean towards.

I am curious what tools you use about which you can say that with such confidence?

Very likely low number of users with a high cost per user. There’s no need for them to expand their market because they’ve got a captive market... no competition. If you’re making money hand over fist ONLY making incremental updates to legacy software, there is zero impetus to make Mac compatible software. That would just eat into your profits with development and support costs.

 

[theorist9]

It's not all QC. Part of the problem was that Catalina officially dropped 32 bit support, which unavoidably broke a lot of software. With every major OS release , a bunch of software stops working, there's a wait of many months for many bits of third-party software to be updated, and there are always some bits of "abandonware" that never make it. Pro users don't like expensive downtime or unscheduled changes, and for anybody doing serious cashy business on their Mac, staying one OS version behind is often the prudent thing to do.

Then, you get a brand new Mac and find that it comes with the latest and greatest operating system version released a few weeks ago and can't be downgraded because the old version doesn't recognise the new hardware - so suddenly you've got to solve all the compatibility problems overnight.

Remember: this isn't just your carefully planned 4-year-or-whatever hardware upgrade cycle: throughout that period hardware can fail/get stolen, new employees join and need kitting out... and suddenly you have to deal with a Catalina in a sea of Mojaves...

Apple's secretive habits aren't particularly helpful to organisations trying to plan their hardware budget... Look at the situation now - the cheapest Macs seem to outperform the more expensive Intel models, but we've only been given the vaguest of notions when the rest of the Intel machines will be replaced or for how long Intel will be supported. Meanwhile for something like, say, audio production, there's a whole laundry list of key 3rd party suppliers who don't yet support Big Sur, let alone M1 and just say "watch this space" on their websites. Logic itself might fly, but whether your virtual instruments, effects, drivers etc. will work properly is a lottery.

...and yes, there's a beta period for every OS, but from what I've seen Apple are liable to make significant changes right up to the final release, and no opportunity for large scale beta testing until the OS is generally available.

You launched into this huge explanation about Catalina without having understood the post to which I was responding. There, the poster mentioned his buddy's apps broke because of a small point update—not an update to a new OS! And since it happened with the Mac Pro, it must have been a point update within Catalina (the Mac Pro is not compatible with earlier OS's, and Big Sur had not yet been released).

Thus I mentioned an instance in which one of Catalina's later point updates broke some apps. That had nothing to do with Catalina's dropping 32-bit support, or any other Catalina-specific change, since these would have been apps that had been happily working in Catalina, that then stopped working with the point update.

 

[theluggage]

You launched into this huge explanation about Catalina without having understood the post to which I was responding.

You're right - I didn't read the GP post properly. Sorry. (I guess that in the spirit of the internet I should try to bluster through with a claim that 10.14 to 10.15 is a point update... but let's skip it ).

Still, forcing OS updates with new/updated hardware is an issue, and the Catalina issues did affect Mac Pro users - and although it's unavoidable with the first round of Apple Silicon & 11.0, we'll probably have MacOS 11.1 "SFO BART station" to contend with when the ASi Mac Pro comes out...

 

[Marinna]

With no VM support, no large memory system (at least 1T DDR4 to start with), both AMD and NVidia out of picture by 2025, no CUDA and OpenCL roadmap, good luck on your server market.
I thought I saw a ray of hope when they introduce the 2019 MAC Pro, now with this new belief Apple plan to go all in-house with their next gen machine, perhaps after all these years Apple still don't care what guys at server market wants.
I am a 16 years Mac user, had at least two Mac Pro including the latest 2019 model. I hope this machine is not going to be the last Mac I ever own when it comes to serious computing.

 

[Yebubbleman]

I'm skeptical that the Mac Pro will be ready before 2022. Bloomberg was quick to assume we'd be getting an Apple Silicon 16" MacBook Pro in 2020. I do believe we'll, at the very least, get the smaller sized iMac and the rest of the Mac mini and MacBook Pro models transitioned over in 2021.

 

[HowardEv]

Then obviously what you want is a Mac Pro or maybe Mac Mini once Apple update the high-end version.

I still need the portability of the Macbook Pro, so 16 cores in the 16" shell is what I'm dreaming for.

Right, what we need is a portable Mac Mini with a battery. The M1 doesn’t have to be plugged in to 120v like ancient desktop computers did. Separate the hot CPU from the screen and keyboard, maximize both CPU performance and laptop thinness and lightness and portability.

 

[Velli]

Right, what we need is a portable Mac Mini with a battery. The M1 doesn’t have to be plugged in to 120v like ancient desktop computers did. Separate the hot CPU from the screen and keyboard, maximize both CPU performance and laptop thinness and lightness and portability.

Sorry dude - no matter how many times you repeat it, I’m pretty sure you are the only customer for that product...

 

[Unregistered 4U]

perhaps after all these years Apple still don't care what guys at server market wants.

No, they don’t and haven’t for awhile. Apple recognizes that to be profitable, they need to focus on a profitable set of customers. That, by default excludes a large number of markets. When Apple had the XServe, it was, for it’s performance priced lower than some of the poorer performing completion, but still made a nice profit for Apple. Once that was no longer true, no further need to go after that market. There may be a few people that are able to USE it like a server, but that’s not a market Apple is targeting.

If by “serious computing” you mean cross-platform Intel based software solutions or libraries, then it very well may be the last Mac Pro you own.

 

[sideshowuniqueuser]

Blackberry also continued to sell sell for a while after the first iPhone was released. The market takes time to react to such stuff, especially when Apple’s full slate of Macs and processors are not out yet.

In the long run though, one bets against Apple to their own detriment. That is one of the few constants in this world.

Au contrare, for heavily traded stock, the price moves almost instantaneously with news. Intel isn't being dumped, and for good reason. Yes, Apple's chips are awesome, but they are only available for Apple products, which are all high end. They also don't run Windows. Intel's price went up recently on news of advances in their quantum computing chips. And so on.

 

[Marinna]

No, they don’t and haven’t for awhile. Apple recognizes that to be profitable, they need to focus on a profitable set of customers. That, by default excludes a large number of markets. When Apple had the XServe, it was, for it’s performance priced lower than some of the poorer performing completion, but still made a nice profit for Apple. Once that was no longer true, no further need to go after that market. There may be a few people that are able to USE it like a server, but that’s not a market Apple is targeting.

If by “serious computing” you mean cross-platform Intel based software solutions or libraries, then it very well may be the last Mac Pro you own.

I agree. then why even bother even sell an Apple silicon powered Mac Pro? iMac/Pro can give them enough profit. This will be especially true when they chip off a large chunk of IT buyers. Small studio design firms won't be able to create enough revenue to even justify the project. A performant but not conformant Mac Pro like the trash can doesn't sell, they should already learned it the hard way last time.

For me, I have been running rather heavy level of photo/video processing and machine learning algorithm on Mac Pro plus at least three VMs running concurrently in the same time. MacOS already lag BIG TIME on the ecosystem of server AI by their narcissistic stubbornness with Metal, and they are kicking out AMD from their ecosystem right now as well. Losing concurrency on Mac Pro will be worse than adding salt to the wound. It has nothing to do with Intel optimized libraries and tools but that they have no offering to ANY real computing but selling to users that only does web-browsing, office software, and dedicated accelerator-enabled multimedia processing.

Some of the YouTuber might said "Mac Pro is good, only at the workflow that Apple is interested" It is not a general purpose machine by many's standard even today. If you ask me, that picture is rather sad and I certainly hope it won't happen. Personally I don't really expect Apple silicon Mac Pro would become a thing until not 2022 but 2025, with another 3~5 years of healthy software support I guess I should still be able to see return of my investment by then.

 

[Unregistered 4U]

then why even bother even sell an Apple silicon powered Mac Pro?

Because there are those that want to, I suppose, do NON-serious, non-real computing, like Logic Pro and Final Cut Pro as fast as they possibly can. If having a complex score or a huge media heavy timeline is an option, they would pay more to make that happen. In some cases, it could be as simple as getting more work done and approved in a day, in others it could be that a producer hates to “freeze” their tracks to avoid overtaxing their processor.

I can’t help but think Apple’s moves over the prior years has been to show to users that if you want the fastest macOS and macOS app performance IN THE WORLD, then we have a high end product for you. For anything else, there’s a world of non macOS choices one of which is SURE to fit your need and price range. They understand very well how large that pool of folks are and most have already moved on due to Apple laggard activity in the high end performance area. Those that are left are either waiting for Apple to turn around, come to their senses, and really, finally, make the system of their dreams they’ve been looking for at the price they’re willing to pay... for REAL this time! OR they’re just so much more productive (or just so much more familiar) with macOS that they’re not going anywhere. They deal with the slow updates and limited upgradability and expensive options because it’s their preference AND more than likely, their pay scale makes it possible for them to say with macOS.

For that group of folks, the coming years holds delight and the fastest macOS performance they’ve ever seen. For others, it’s likely finally, at long last, time to start that painful move to some other OS. I have little doubt that, while painful in the short run, most of those folks will end up giddy with the options, upgradability, and raw performance available to them.

 

[Alan Wynn]

Actual real software (like the stuff used to design the Apple Macbook PCB):
This is only availabe for Windows - no chance of Windows for ARM versions:

If the higher end Apple Silicon Macintosh systems are as far ahead of the Intel/AMD higher end systems as these current M1 based systems are against their competitors, I would not be surprised to see a Windows on Arm port as the easiest way to support Apple Silicon. Baring that, from what we have seen, running translated will still out perform native x86_64 hardware.

Altium Designer (6500 euro / seat)
Tektronix SignalVu (needed to operate my 10 000 USD real-time spectrum analyser)
Zuken CadStar
Sonnet Suite EM (unless you buy the Gold version, which has RedHat Support - but not redHat on ARM)
Various other circuit simulators

I'm often at client sites - so a laptop is necessary - don't want to use 2 laptops and prefer Mac side of things for file management, admin workflow etc.

For all of these, it will be a question as to how great a performance lead they have. It it is substantial, we will see ports to ARM, and if it is very substantial, we might even see native macOS versions.