Impressive but unbelievably expensive considering what will actually be run on it.
I kind of want one, but the 5800x for example is about 10x less expensive than a specced out MacBook Pro
I kind of want one, but the 5800x for example is about 10x less expensive than a specced out MacBook Pro
All of Apple's innovation seems to be towards better and cheaper AR/VR hardware. Desktop-class processors would be a distraction for them.
And with all of the top cloud players building custom silicon these days, there is little room for Apple to sell CPUs to the server market even if they were inclined (which they are not).
The only strategic Apple vertical that might align with desktop-class CPUs is the Apple Car initiative and specifically self-driving. Dedicated audio/image/video processing and inference focused hardware could better empower creatives for things like VFX or post-processing or media in general. However, its not clear to me that is enough of a market for Apple's unit economics compared with more iDevice / MacBook sales.
What's the alternative? That they release another Mac Pro with intel, despite their stated intention to move everything away from x86, or that they release a Mac Pro with just a laptop chip inside?
Let's remember that Apple has an annual R&D budget of c$20 billion, so it won't be totally shocking if they diverted a small fraction of that to build out a desktop processor.
Simply quadrupling the die isn't really feasible, the price increases exponentially.
Even if the SOC was $500 cost price they would still have plenty of room to pay for everything else and hit the $5k price point.
I mean, they could even use separate highly integrated chips to get that number of cores too, although I suspect they will want to do it on a single chip assuming cooling is OK.
What you're talking about at the end is chiplets and it takes a lot of work to get it to operate, just ask AMD.
I guess the question is, to what end?
- Apple hasn't had a good history with B2B sales. Evidenced by them getting out of the server business.
- Apple doesn't want to be a cloud infra company. Evidenced by them using other cloud providers to host iCloud.
- Maybe, Apple will one day sell servers again, or sell M1-like chips to cloud providers. However, we have no evidence of this. Apple prefers B2C products with 5-6 year refresh cycles. Their superpower is vertical integration rather than selling a single component.
- Cloudtops are becoming more and more popular, in-tandem with a powerful-enough laptop to appropriately hold local state such that the UX is responsive. Large companies that do software development, ML training, VFX, etc are already using cloudtops.
That said, small companies and individuals haven't yet started using cloudtops for intellectual property creation and a Mac Pro could still help their workflows.
With that in mind, where does Apple grow this high performance market? If there is no future for this market, why invest into any R&D? Let the Windows ecosystem continue to handle it for the few years it still exists.
Meanwhile, to help those small companies and individuals improve their current workflows, maybe Apple will release one or two more Mac Pros, but more likely than not all Mac Pro development will stop. I'm likely wrong about all of this but I feel, Apple will instead offer a "seamless compute cloud". Similar to their strategy with https://developer.apple.com/xcode-cloud/ they will likely release a general purpose service for "cloud cores" i.e. the machine will locally have 6 high-performance cores, 2 high-efficiency cores, and ∞ remote-performance cores that are seamlessly integrated and run Apple's Universal Binary in "their cloud" (likely through AWS/Azure/GCP). Then no matter what the workflow is (e.g. photo/video/audio/game editing) you get a local preview (e.g. a low-res render locally) but in parallel the final product is built remotely and the data mirrored back to your local machine.
The only other reasons to keep building beefy machines like Mac Pro are for iCar Self Driving or to sell "cloudtops" (as I described in the paragraph earlier) but in a B2C context, i.e. the cloudtop runs in your local LAN/WAN or even as your home's Wifi router to reduce latency and accelerate all the Apple devices on the network. However, its not clear if they will invest in the software first or the hardware first for this distributed compute future.
They have made a Mac Pro desktop for several decades. I am trying to follow your reasoning for Apple to sunset that category of workstation as a result of transitioning to Apple silicon, but it is not working out for me.
My logic leads to a cheaper-to-produce-than-Macbook workstation in an updated Mac Pro chassis with best-binned M1X Max parts in the Spring followed by its first chiplet-based Apple silicon workstation using the same chassis in the Fall, followed by an annual iteration on Apple silicon across its product line, on about the same cadence as A/A-x iOS devices.
Part of my reasoning is based on the assumption that Mac Pro sales generate more Pro XDR sales displays at a higher rate than Macbook Pro sales. I think the total profit baked into an Apple silicon Mac Pro + Pro XDR is big and fills a niche not filled by any thing else in the market. Why leave it unfilled?
Yeah this seems realistic to me. Specifically, I meant that there will be no D1 (desktop-class chip) to complement the A14/A15 (mobile-class chips) and M1/Pro/Max (laptop-class chips).
I agree that taking M1 Max, or future laptop-class chips (e.g. M2 Max) and adding better cooling and perhaps a chiplet would be most likely.
That said, Apple's R&D is always for some grand future product vertical, not to make some more money over the next few quarters. e.g. Neural Engine and Security Enclave R&D leads 1. to on-device Siri + Face Id to accommodate a marketing change to focus on Privacy; 2. moves towards a password-less future.
Even R&D for chiplets, may not make sense for them other than for Apple Cars. "Can you explain your unit economic analysis?" during a chip shortage, why sell 1 Mac Pro with a 10+ year "lifespan" when Apple could sell 5 iDevices with a 5+ year "lifespan" instead? e.g. had Apple known better that they would need to reduce iPhone production targets because of chips, its likely they would have delayed the M1 Pro/Max production/launch.
The issue to me is that because apple is the only one who can put these in computers this will have no real effect on PC component pricing. Apple makes, apple puts in apple computers, can't run windows, if you're in another ecosystem it's not even a real option.
Still might buy one, I want an arm box to test my work on compilers on.
My current machine has like 10 billion (GPU + CPU)
I suppose it’s not likely we’ll know the actual price of the M1, but it would suffice to say it’s probably a fair bit less than the full laptop.
https://www.tomshardware.com/news/tsmcs-wafer-prices-reveale...
Since the M1 is 120 mm2 and a 300mm wafer is about 70695 mm2, you could theoretically fit 589 M1 chips on a wafer.
Subtract the chips lost at the edge and other flaws, you might be able to get 500 M1 off a wafer? (I know nothing about what would be a reasonable yield but I'm pretty sure a chip won't work if part of it is missing)
Anyways, that would be $17000/500, or $34 per M1 chip - based just on area and ignoring other processing costs.
TSMC has stated that N7 defect density was 0.09 a year ago or so. They have also since stated that N5 defect density was lower than N7.
Let's plug that in here https://caly-technologies.com/die-yield-calculator/
If we go with a defect density of 0.07, that's 94 good dies and 32 harvest dies. At 0.08, it's 90 and 36 respectively.
If we put that at 120 dies per wafer and $17,000 per wafer, that's just $141 per chip. That's probably WAAAYY less than they are paying for the i9 chips in their 2019 Macbook Pros.
For comparison, AMD's 6900 GPU die is 519mm^2 and Nvidia's 3080 is 628mm^2. Both are massively more expensive to produce.
And I agree - Apple making their own chips is sure looking good for their bottom line vs buying from Intel
I also don't need portability nor a fancy display because I use all external monitors and keyboard/mouse. Fan noise is irrelevant because I wear noise cancelling headphones.
For me, it's not worth spending the $2500 on a MacBook when I can get similar/better performance with a much cheaper upgrade.
It's hard to compare to laptops, but since we started the desktop comparison, a Ryzen 7 5800X is $450 MSRP, or about $370 current street price. Motherboards can be found for $100, but you'll more likely spend $220-250 for a good match. 32GB RAM is $150, 1TB Samsung 980 Pro is a bit under $200. Let's assume desktop RTX 3060 for the graphics (which is probably slightly more powerful than the 32-core GPU M1 Max) for MSRP $330 but street price over $700.
So we're at about $1670 for the components, before adding in case ($100), power supply ($100) and other things a laptop includes (screen, keyboard...).
In my case I'm reusing everything and just upgrading the CPU for $560. The end result will be similar or better performance than the MacBook.
If you don't own any desktop computer parts at all then, yes, one can easily spend $2500 building a machine with similar performance.