Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Apple M1 Ultra

apple.com

Post reply on HN
Print view View original

511–520 of 855 posts

Re: Apple M1 Ultra

#511

Post by MBCook »

paulmd wrote: post #286
zitterbewegung wrote: post #54

Earlier quoted context omitted.

They might have to have the unified memory more dense to get to 1.5 TB max of RAM on the machine (also since this would be originally shared with a GPU). Maybe they could stack the RAM on the SoC or just get the RAM at a lower process node.

The M1 Max/Ultra is already extremely dense design for that approach, it's really almost as dense as you can make it. There's packages stacked on top, and around, etc. I guess you could put more memory on the backside but that's not going to do more than double it, assuming it even has the pinout for that (let's say you could run it in clamshell mode like GDDR, no idea if that's actually possible, but just hypothetic…

Do people really need 1.5 TB of unified memory? If you had 128 GB of unified memory and another pool of 1.5 TB or whatever of “slower” memory (more like normal speed on the Intel/AMD side) would that work?

You (or the OS or the chip) could page things in and out if the unified memory. Treat unified memory as a MEGA L3 cache.

Depending on how it’s done it may not be transparent if you want the best performance. But would it work?

Re: Apple M1 Ultra

#512

Post by MaxLeiter »

Flankk wrote: post #472
MaxLeiter wrote: post #444

Earlier quoted context omitted.

I don’t think this will be the case until they’re more right to repair friendly

Most people don't want to repair, they want to replace. What people really want is a modular computer, the way things used to be. Right to repair is going to force legislation for schematics, at best. Even then I don't think it's going to happen. Like it or not, Apple builds integrated products now. You are the old man grumbling about the iMac not including a floppy drive.

I'm not complaining; I generally agree with you. But (I believe) for people to invest in Macs for servers they need to be able to fix and replace them faster than dealing with Apple would allow.

Re: Apple M1 Ultra

#513

Post by ricardobeat »

itstomkent wrote: post #502
ricardobeat wrote: post #424

Earlier quoted context omitted.

The vast majority of developers today has a laptop as their main machine. Performance-per-watt is absolutely crucial there.

I'll just never understand this. Chances are you are at the same desk day in and day out. You probably have monitors and external keyboard/mouse hooked up because hunching over a laptop and using a touchpad is unnecessary torture for a fixed workspace. Given that why would you hamstring yourself with a thermally constrained and overpriced-because-miniaturization-isn't free setup? Until maybe these M1's (and I'm not e…

Meetings, working from home, travel, changing desks every few months. Not having a jungle of cables. Retina screen (still hard to find on external monitors). The vast majority of devs also don’t need an insanely powerful machine. If Docker on Mac wasn’t dog slow, I could easily get by with a $999 MacBook Air.

Re: Apple M1 Ultra

#514

Post by mehmetoguzderin »

apohn wrote: post #495
mehmetoguzderin wrote: post #488

Earlier quoted context omitted.

> The 3090 also can do fp16 and the M1 series only supports fp32 Apple Silicon (including base M1) actually has great FP16 support at the hardware level, including conversions. So it is wrong to say it only supports FP32.

Thanks. At least when I ran the benchmarks with Tensorflow, using mixed precision resulted in the CPU being used for training instead of the GPU on the M1 Pro. So if the hardware is there for fp16 and they will implement the software support for DL frameworks, that will be great.

Yes, unfortunately, the software is to blame for the time being, and I also ran into issues myself. :\ Hope they catch up to what the hardware delivers well, including both the GPU and the Neural Engine.

Re: Apple M1 Ultra

#516

Post by Sporktacular »

greenknight wrote: post #222
flatiron wrote: post #181

Earlier quoted context omitted.

People care about performance per watt now. So they could compete. The real question is if they would support Linux. In our containerized world I can’t see their servers getting super big running macOS

The reason they dominate at PPW, is because they are on TSMCs 5nm process. No one else has made a cpu chip on this process yet. AMD are scheduled for later this year (they are currently using 7nm). It will be interesting to see the difference in performance and performance per watt, when both companies are on the same node.

No, it's hetrogeneity, system integration, pipeline width and instruction set as well.

Re: Apple M1 Ultra

#517

Post by EugeneOZ »

kitsunesoba wrote: post #486
OOPMan wrote: post #461

Earlier quoted context omitted.

That would not help. The reason proton works so well is because Linux on x86 and Windows on x86....are both on x86. Proton on ARM Mac's would involve Rosetta and while that does a surprisingly good job of running x86 on ARM I'm not sure it's up to the job of running games at high speed...

Games seem playable running through Microsoft’s inferior x86 compatibility layer in a Windows ARM VM on M1 Pro/Max, so I don’t see how running games through Rosetta would be any worse.

If you have some monster like 3090, you expect it to run games smoothly on “ultra” settings. In reality, some games are pushing its limits even on “middle” settings (Cyberpunk 2077 @ 2k, RDR2 @ 4k). Any emulator will hit performance, so if you want some top-of-the-top performance - you still need Windows PC with RTX 3090. And it will not be even more expensive than Mac Studio ;)

Re: Apple M1 Ultra

#518

Post by MBCook »

MaxLeiter wrote: post #512
Flankk wrote: post #472

Earlier quoted context omitted.

Most people don't want to repair, they want to replace. What people really want is a modular computer, the way things used to be. Right to repair is going to force legislation for schematics, at best. Even then I don't think it's going to happen. Like it or not, Apple builds integrated products now. You are the old man grumbling about the iMac not including a floppy drive.

I'm not complaining; I generally agree with you. But (I believe) for people to invest in Macs for servers they need to be able to fix and replace them faster than dealing with Apple would allow.

Honestly CPUs and memory fail very rarely. If the storage was on standard m.2 sticks and maybe the I/O was on a replaceable module in case Ethernet fails… would that be enough to assuage most server concerns?

I don’t see Apple getting back into that business. But I think they have the ability to make a good option if they want.

Re: Apple M1 Ultra

#519

Post by avl999 »

klelatti wrote: post #7

I wonder if the max 128GB graphics memory opens up some applications that would not have been viable before?

Yep, finally Electron apps will run smoothly on consumer hardware.

Re: Apple M1 Ultra

#520

Post by modzu »

nextos wrote: post #231
mhh__ wrote: post #215

Earlier quoted context omitted.

How many threads are actually getting utilized in those kernel builds? I don't work on the kernel enough to have intuition in mind but people make wildly optimistic assumptions about how compilation stresses processors. Also 1st gen threadrippers are getting on a bit now, surely. It's a ~6 year old microarchitecture.

Yes, it'd be interesting to see this comparison made with current AMD CPUs and a full build that has approximately the same price. I am curious whether there is a real performance difference? I do lots of computing on high-end workstations. Intel builds used to be extremely expensive if you required ECC. They used that to discriminate prices. Recent AMD offerings helped enormously. I wonder whether these M1 offerings…

but do the m1s have ecc?
Post reply on HN