M1 Apple Silicon for Chess?

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
Milos
Posts: 4025
Joined: Wed Nov 25, 2009 12:47 am

Re: M1 Apple Silicon for Chess?

Post by Milos » Mon Dec 07, 2020 8:40 am

MRI_Doc wrote:
Mon Dec 07, 2020 5:38 am
You get max performance with 4 threads, which utilizes only the high performance cores.

Adding more threads results in SLOWER calculation, as the code must not be optimized for variable speed cores.

Run your benchmarks with 4 threads and the M1 performs very well.
That is simply a ridiculous argument, or at least the underlying reasoning. Slower cores run exactly the same instruction set as faster cores just slower. SF code is agnostic to individual thread speed and more threads will always give you higher nps, no matter what. It they didn't there would be some weird blocking condition in parallel execution indicating a bug in OS scheduler which is highly unlikely.
What you are probably observing is simply that with all 8 cores running, there is some thermal throttling which is not the case with 4 cores running.

twobeer
Posts: 20
Joined: Sat Feb 22, 2020 5:43 pm
Full name: Leif Aronsson

Re: M1 Apple Silicon for Chess?

Post by twobeer » Mon Dec 07, 2020 5:02 pm

These benchmark makes me wish more for x86 5nm process chips arriving faster rather than some more overhyped Apple ARM chips (or QUALCOMM's for that matter) placed in proprietary, overpriced, service/expansion-unfriendly, locked up and locked in devices.

For the same price of a thermally challenged Macbook Air M1 we can get an Asus TUF 15.6" Gaming Laptop with AMD Ryzen 9 4900H/1TB SSD/16GB RAM/GeForce RTX 2060 , 144hz screen, 90W battery)+ expandable RAM to 32GB, and extra ssd slots) ..

George Sobala
Posts: 28
Joined: Sat Feb 03, 2018 1:42 pm
Location: Yorkshire, England

Re: M1 Apple Silicon for Chess?

Post by George Sobala » Mon Dec 07, 2020 5:14 pm

... which will have blown up on you by 2-3 years.

mmt
Posts: 343
Joined: Sun Aug 25, 2019 6:33 am
Full name: .

Re: M1 Apple Silicon for Chess?

Post by mmt » Mon Dec 07, 2020 10:30 pm

Apple Preps Next Mac Chips With Aim to Outclass Top-End PCs
https://www.bloomberg.com/news/articles ... st-end-pcs

syzygy
Posts: 4999
Joined: Tue Feb 28, 2012 10:56 pm

Re: M1 Apple Silicon for Chess?

Post by syzygy » Mon Dec 07, 2020 11:29 pm

twobeer wrote:
Mon Dec 07, 2020 5:02 pm
These benchmark makes me wish more for x86 5nm process chips arriving faster rather than some more overhyped Apple ARM chips (or QUALCOMM's for that matter) placed in proprietary, overpriced, service/expansion-unfriendly, locked up and locked in devices.
The Apple ARM chips were underhyped. They beat all expectations.
Do you think one or two x86 die shrinks will give 10x better performance per Watt?

If you buy Apple, you obviously pay for the brand name, but you also pay for a build quality that you won't find in an Asus gaming laptop. Apple laptops aren't much more expensive than Dell XPS laptops.

I won't be buying an Apple M1 laptop though, as I wouldn't be able to run Linux on it. But an Apple M1 Mac mini would make an interesting toy.

User avatar
jshriver
Posts: 1242
Joined: Wed Mar 08, 2006 8:41 pm
Location: Morgantown, WV, USA
Contact:

Re: M1 Apple Silicon for Chess?

Post by jshriver » Mon Dec 07, 2020 11:59 pm

After reading I'm interested. The mac mini for under $700 seems like a nice price and cheaper than some high end graphics cards alone.

My poor wallet this time of year lol.

Alayan
Posts: 531
Joined: Tue Nov 19, 2019 7:48 pm
Full name: Alayan Feh

Re: M1 Apple Silicon for Chess?

Post by Alayan » Tue Dec 08, 2020 12:14 am

syzygy wrote:
Mon Dec 07, 2020 11:29 pm
The Apple ARM chips were underhyped. They beat all expectations.
Do you think one or two x86 die shrinks will give 10x better performance per Watt?

If you buy Apple, you obviously pay for the brand name, but you also pay for a build quality that you won't find in an Asus gaming laptop. Apple laptops aren't much more expensive than Dell XPS laptops.

I won't be buying an Apple M1 laptop though, as I wouldn't be able to run Linux on it. But an Apple M1 Mac mini would make an interesting toy.
Even Apple apologists don't claim more than the 3-4x perf-per-watt that you get when comparing a mobile M1 designed for low power consumption with a stock desktop x86 CPU designed for max performance. 10x is fantasy.

And this way of comparing perf-per-watt is completely wrong when it comes to architectural comparisons.

This is like when Nvidia compared an undervolted and underclocked RTX 3080 to a RTX 2080 at iso-performance, putting them at very different points of their efficiency curve, to claim a 2x perf-per-watt advantage for Ampere compared to Turing. The real improvement from architectural changes and process nodes was somewhere around 20% (GPUs going wider make it hard to normalize).

The proper measure of performance per watt is performance at iso-wattage. A laptop chip or a server chip, for example, aren't designed around a target level of performance with then an adjusted wattage. They are designed around a target wattage and try to get the highest level of performance possible within these power consumption limit, which often involves finding ways to reduce power consumption.

This is even more relevant for CPUs where normalization is easy by looking at single-core performance. Comparison at iso-wattage is much more resilient to the distortions of perf/watt introduced by the efficiency curve. Increasing clocks increases power consumption by itself, but also increases the required voltage increasing power consumption once again. Chasing the last few percents of performance make power consumption explode.

A Zen 3 core with reduced clocks and voltage to use similar power to a M1 core ends up having 20-25% less performance (25-33% perf advantage for the M1). A sizable advantage for Apple, but not a mindblowing difference that would condemn x86 CPUs to irrelevance. A lot of Apple's advantage is the fruit of them hiring a lot of very highly skilled engineers and giving them the transistor budget to do something great (Apple designs are not focused on minimizing die space), this is not by itself a proof that ARM is significantly better than x86.

It would be tempting to correct for Apple's 5nm vs 7nm node advantage by applying the TSMC claim of 30% power reduction to the Zen3 core, getting a similar perf/watt, but that would be a mistake because it would break the iso-power comparison. The performance of a Zen3 core with a third more power than a M1 core is a better comparison, and the Zen3 core still loses, but by a thinner margin than when not correcting for process node.

You'll notice that Apple's own Icestorm cores have a better perf/watt than their Firestorm cores, yet the hype comes from the big Firestorm cores. This is also a reminder that the absolute level of performance reached is, for many workloads, very important. The sole perf/watt is not the only relevant factor for a responsiveness-oriented workload rather than an instance-parallelizable throughput-oriented workload like many server tasks are.

If one designs peaks at 1.00 perf at 5W but can't reach more performance at 15W, and another design gets 0.9 perf at 5W but peaks at 1.1 perf at 15W, the first design is the better choice if you have a 5W budget but the second one is better if you have a 15W budget.

Raphexon
Posts: 369
Joined: Sun Mar 17, 2019 11:00 am
Full name: Henk Drost

Re: M1 Apple Silicon for Chess?

Post by Raphexon » Tue Dec 08, 2020 9:06 am

Chip with trash for IO consumes less energy than a big x570 board that has a ton of IO who would have thought?
Once they start adding IO energy consumption will rise with it.
Not because ARM is inherently superior but because Apple can take the cost of making their systems efficient.

Apple doesn't have to bother with chiplets either, they'll just eat the extra cost for some performance gain.
Apple bet on super wide cores being worth it. (which ended up being the case)
I guess their IO die is on 5nm too, while Zen3 is still on 14nm for cost-related reasons.

So no superiority of ARM, just Apple having the engineers and money to make it work.

Also if the Nvidia-ARM deal doesn't get blocked, the future is RISC-V.
Future might be Risc-V regardless but Nvidia-ARM will definitely make risc-v more attractive.

smatovic
Posts: 1648
Joined: Wed Mar 10, 2010 9:18 pm
Location: Hamburg, Germany
Full name: Srdja Matovic
Contact:

Re: M1 Apple Silicon for Chess?

Post by smatovic » Tue Dec 08, 2020 11:17 am

Raphexon wrote:
Tue Dec 08, 2020 9:06 am
...
Also if the Nvidia-ARM deal doesn't get blocked, the future is RISC-V.
Future might be Risc-V regardless but Nvidia-ARM will definitely make risc-v more attractive.
I agree, on one side their is a lockdown of Apple ARM and who knows what will happen with ARM+Nvidia, so on the other side is room and demand for an open chip arch like RISC-V, SiFive released recenty IP cores with RISC-V Vector ISA, so the arch seems to be on track...

--
Srdja

syzygy
Posts: 4999
Joined: Tue Feb 28, 2012 10:56 pm

Re: M1 Apple Silicon for Chess?

Post by syzygy » Tue Dec 08, 2020 10:16 pm

Raphexon wrote:
Tue Dec 08, 2020 9:06 am
Chip with trash for IO consumes less energy than a big x570 board that has a ton of IO who would have thought?
Once they start adding IO energy consumption will rise with it.
Not because ARM is inherently superior but because Apple can take the cost of making their systems efficient.

Apple doesn't have to bother with chiplets either, they'll just eat the extra cost for some performance gain.
Apple bet on super wide cores being worth it. (which ended up being the case)
I guess their IO die is on 5nm too, while Zen3 is still on 14nm for cost-related reasons.

So no superiority of ARM, just Apple having the engineers and money to make it work.

Also if the Nvidia-ARM deal doesn't get blocked, the future is RISC-V.
Future might be Risc-V regardless but Nvidia-ARM will definitely make risc-v more attractive.
So you agree that x86 is going the way of the light bulb.

Post Reply