Apple M1 Max Geekbench Score • Hacker News Board

Re: Apple M1 Max Geekbench Score

#581

Post by b9a2cab5 » Thu, Oct 21, 2021, 4:58 AM UTC

shaklee3 wrote: post #579

lanza wrote: post #558

Earlier quoted context omitted.

You're missing out on the fact that Apple didn't release a 12980hk or 5980hx competitor. These are 30 watt chips that trounce the competition's 65 watt (e.g. the 12980hk and 5980hx) and beyond chips. Hell, this Geekbench is faster than a desktop 125 watt 11900k. It's faster than a desktop 105 watt 5800x. Apple intentionally played to the competition here. They know AMD/Intel reach some performance level X and release…

You are vastly, vastly overestimating how many people like apple laptops. They're expensive, the OS is worse, and there's no speed difference outside of benchmarks. That combined with the fact that most apps are windows-based, and the games are poorly emulated or nonexistent, there's really no reason to buy an apple laptop short of bragging.

Is that why every tech company uses Macbooks?

People who can't afford Apple laptops (and gamers and certain other use cases) buy cheaper laptops. The brand appeal draws all the wealthy people towards Apple devices.

Re: Apple M1 Max Geekbench Score

#582

Post by shaklee3 » Thu, Oct 21, 2021, 4:58 AM UTC

derefr wrote: post #410

jjcon wrote: post #284

Earlier quoted context omitted.

> GPU capabilities that blow everything away Compared to previous macs and igpus - an nvidia gpu will still run circles arounnd this thing

> an nvidia gpu will still run circles arounnd this thing Not for loading up models larger than 32GB it wouldn't. (They exist! That's what the "full-detail model of the starship Enterprise" thing in the keynote was about.) Remember that on any computer without unified memory, you can only load a scene the size of the GPU's VRAM. No matter how much main memory you have to swap against, no matter how many GPUs you thro…

Terrible, terrible take. GPU ram is about 10x faster than ddr in practical usage, so your worloads can finish and a new batch can be transferred over pcie faster than the apple would finish the first pass.

Re: Apple M1 Max Geekbench Score

#583

Post by noahtallen » Thu, Oct 21, 2021, 5:00 AM UTC

soylentnewsorg wrote: post #569

imwillofficial wrote: post #568

Earlier quoted context omitted.

I mean sure, compare your 25lbs Lenovo P52 with a 3lb MacBook Air. Not really apples to apples there bud.

I love how you read the first line, ignore the rest of the paragraph addressing your specific point, then proudly strut around like you called out a gotcha. "bud." Dell Precision. 6 core xeon, 128GB RAM, 9 hour battery life. Thin and under 7lb. Metal and solid, can run over it w/ a car just like the Lenovo you're talking about. Has had a 4k screen and keyboards that don't break or imprint on the display. 5 years ago.…

You probably wouldn’t have gotten downvoted as much if people hadn’t (ironically?) read the part you said they wouldn’t read :p And I want to mention that I agree with you about apple not selling to pros very well over the past 6 years.

Your laptop is definitely very capable. But it’s barely a laptop. Why not build out a proper desktop? This precision would be a pain in the ass to carry around for most people, especially travel. Dell made sacrifices to get that kind of power: size, cooling, and battery life. Those are actually meaningful things when it comes to a laptop for most people, even pros.

I think the fact that you’re even mentioning a MacBook air in the same sentence is very good for the Air. The M1 hasn’t really been marketed to the pro market until the recent release.

Also, 5x faster at what? The M1 is about the same performance as a W-11855M at single core cinebench, and only 25% slower at multicore. So comparison to the M1 pro/max is not very promising for the Xeon chip.

Re: Apple M1 Max Geekbench Score

#584

Post by shaklee3 » Thu, Oct 21, 2021, 5:00 AM UTC

derefr wrote: post #433

jjcon wrote: post #418

Earlier quoted context omitted.

Memory != Speed You could throw a TB of memory in something and it won't get any faster or be of any use for 99.99% of use cases. Large ML architectures don't need more memory, they need distributed processing. Ignoring memory requirements, GPT-3 would take hundreds of years to train on a single high end GPU (on say a desktop 3090 which is >10x faster than m1) which is why they aren't trained that way (and why NVidia…

I don't disagree that there are many use cases for which more memory has diminishing returns. But I would disagree that those encompass 99.99% of use cases. Not all problems are embarrassingly-parallel. In fact, most problems aren't embarrassingly parallel. It's just that we mostly use GPUs for embarrassingly-parallel problems, because that's mostly what they're good at, and humans aren't clever enough by half to com…

Again, there is a huge memory bandwidth bottleneck. It's ddr versus gddr and hbm. It's not even close. The m1 will be slower.

Re: Apple M1 Max Geekbench Score

#585

Post by shaklee3 » Thu, Oct 21, 2021, 5:01 AM UTC

b9a2cab5 wrote: post #581

shaklee3 wrote: post #579

Earlier quoted context omitted.

You are vastly, vastly overestimating how many people like apple laptops. They're expensive, the OS is worse, and there's no speed difference outside of benchmarks. That combined with the fact that most apps are windows-based, and the games are poorly emulated or nonexistent, there's really no reason to buy an apple laptop short of bragging.

Is that why every tech company uses Macbooks? People who can't afford Apple laptops (and gamers and certain other use cases) buy cheaper laptops. The brand appeal draws all the wealthy people towards Apple devices.

I've worked at many tech companies and apple are by far the minority. And that's a biased sample size of really the largest market they have.

Re: Apple M1 Max Geekbench Score

#586

Post by b9a2cab5 » Thu, Oct 21, 2021, 5:01 AM UTC

soylentnewsorg wrote: post #569

imwillofficial wrote: post #568

Earlier quoted context omitted.

I mean sure, compare your 25lbs Lenovo P52 with a 3lb MacBook Air. Not really apples to apples there bud.

I love how you read the first line, ignore the rest of the paragraph addressing your specific point, then proudly strut around like you called out a gotcha. "bud." Dell Precision. 6 core xeon, 128GB RAM, 9 hour battery life. Thin and under 7lb. Metal and solid, can run over it w/ a car just like the Lenovo you're talking about. Has had a 4k screen and keyboards that don't break or imprint on the display. 5 years ago.…

I guarantee you, an M1 is about as fast as your "6 core xeon" laptop. M1 Pro/Max will steamroll it. You can look at Cinebench, LLVM compile, Java benchmarks, etc. You're completely delusional claiming your laptop is 5x faster in CPU. Mocking a "phone CPU" when the A15 is actually faster than a 5950X in single core performance shows you don't know what you're talking about.

Re: Apple M1 Max Geekbench Score

#587

Post by dijit » Thu, Oct 21, 2021, 5:03 AM UTC

soylentnewsorg wrote: post #569

imwillofficial wrote: post #568

Earlier quoted context omitted.

I mean sure, compare your 25lbs Lenovo P52 with a 3lb MacBook Air. Not really apples to apples there bud.

I love how you read the first line, ignore the rest of the paragraph addressing your specific point, then proudly strut around like you called out a gotcha. "bud." Dell Precision. 6 core xeon, 128GB RAM, 9 hour battery life. Thin and under 7lb. Metal and solid, can run over it w/ a car just like the Lenovo you're talking about. Has had a 4k screen and keyboards that don't break or imprint on the display. 5 years ago.…

I also have a Xeon laptop. (45w TDP E3-1505m v6 in a dell precision).

Xeons are not magically faster than their i7/i9 counterparts (mine being not faster than a i7-7820HQ which is its contemporary flagship high performance mobile CPU). In fact they can be slower because the emphasis is on correctness and multi core, not usually single thread performance.

Xeons are also slower than a modern AMD chip which also can have more cores.

5x is a performance metric that doesn’t match up. Unless you have a desktop class 145w/165w cpu, in which case it’s not going to get 9hrs of battery unless you’re not actually touching the CPU. More like 30 minutes of heavy use on the largest battery legally allowed in laptops.

Edit: I just took a quick snoop on geek bench and found a modern xeon w in the largest dell precision laptop available:

Single core: 929 vs 1783 for M1

Multi core: 6718 vs 12693 for M1

https://browser.geekbench.com/v5/cpu/10517773

Synthetic scores aren’t everything, but I’m hard pressed seeing how you can get 5x performance out of a chip that scores almost exactly half. Even with hardware accelerations like AVX512 (which causes a cpu to get so hot it throttles below baseline even on servers with more than adequate cooling.)

Re: Apple M1 Max Geekbench Score

#588

Post by eyelidlessness » Thu, Oct 21, 2021, 5:12 AM UTC

Someone wrote: post #263

ur-whale wrote: post #93

Earlier quoted context omitted.

> Not when building large software projects. Or run heavy renders of complex ray-traced scenes. Or do heavy 3D reconstruction from 2D images. Or run Monte-Carlo simulations to compute complex likelihoods on parametric trading models. Or train ML models. The list of things you can do with a computer with many, many cores is long, and some of these (or parts thereof) are sometimes rather annoying to map to a GPU.

It seems Apple thinks it _can_ map the essential ones to the GPU, though. If they didn’t, there would be more CPUs and less powerful other hardware. ‘Rather annoying’ certainly doesn’t have to be a problem. Apple can afford to pay engineers lots of money to write libraries that do that for you. The only problem I see is that Apple might (and likely will) disagree with some of their potential customers about what func…

Re: Apple M1 Max Geekbench Score

#589

Post by Hamuko » Thu, Oct 21, 2021, 5:13 AM UTC

dcolkitt wrote: post #466

blakesterz wrote: post #100

I just can't figure out what I'm missing on the "M1 is so fast" side of things. For years I worked* on an Ubuntu desktop machine I built myself. Early this year I switched to a brand new M1 mini and this this is slower and less reliable than the thing I built myself that runs Ubuntu. My Ubuntu machine had a few little issues every no and then. My Mini has weird bugs all the time. e.g. Green Screen Crashes when I have…

MacOS just seems to have a lot of stability problems related to peripherals. I don’t think it’s specific to M1. Even on my late-era Intel MBP, Bluetooth and USB often triggers a kernel panic.

When I got my first MacBook Pro, there was actually bug in iPhone USB Tethering that caused a guaranteed kernel panic if you tried to do it. Managed to crash my laptop a few times with my iPhone 3GS.

Re: Apple M1 Max Geekbench Score

#590

Post by darksaints » Thu, Oct 21, 2021, 5:14 AM UTC

b9a2cab5 wrote: post #581

shaklee3 wrote: post #579

Earlier quoted context omitted.

You are vastly, vastly overestimating how many people like apple laptops. They're expensive, the OS is worse, and there's no speed difference outside of benchmarks. That combined with the fact that most apps are windows-based, and the games are poorly emulated or nonexistent, there's really no reason to buy an apple laptop short of bragging.

Is that why every tech company uses Macbooks? People who can't afford Apple laptops (and gamers and certain other use cases) buy cheaper laptops. The brand appeal draws all the wealthy people towards Apple devices.

I work at one of those tech companies, and have been issued a brand new $4000 MacBook pro. It sucks. Constantly crashing, requires multiple restarts to get networking to function properly if the battery dies while in sleep mode, and it has an extremely slow file system even with the best SSD they have available, and there are always 30 different things that break whenever I have to install OS updates.

Tech companies buy Apple products because they try to cater to expensive tech workers, and there is a massive subset of tech workers that are brand-obsessed culty shitnoggins.