Apple's M1 Pro, M1 Max SoCs Investigated: New Performance and Efficiency Heights
by Andrei Frumusanu on October 25, 2021 9:00 AM EST- Posted in
- Laptops
- Apple
- MacBook
- Apple M1 Pro
- Apple M1 Max
CPU ST Performance: Not Much Change from M1
Apple didn’t talk much about core performance of the new M1 Pro and Max, and this is likely because it hasn’t really changed all that much compared to the M1. We’re still seeing the same Firestrom performance cores, and they’re still clocked at 3.23GHz. The new chip has more caches, and more DRAM bandwidth, but under ST scenarios we’re not expecting large differences.
When we first tested the M1 last year, we had compiled SPEC under Apple’s Xcode compiler, and we lacked a Fortran compiler. We’ve moved onto a vanilla LLVM11 toolchain and making use of GFortran (GCC11) for the numbers published here, allowing us more apple-to-apples comparisons. The figures don’t change much for the C/C++ workloads, but we get a more complete set of figures for the suite due to the Fortran workloads. We keep flags very simple at just “-Ofast” and nothing else.
In SPECint2017, the differences to the M1 are small. 523.xalancbmk is showcasing a large performance improvement, however I don’t think this is due to changes on the chip, but rather a change in Apple’s memory allocator in macOS 12. Unfortunately, we no longer have an M1 device available to us, so these are still older figures from earlier in the year on macOS 11.
Against the competition, the M1 Max either has a significant performance lead, or is able to at least reach parity with the best AMD and Intel have to offer. The chip however doesn’t change the landscape all too much.
SPECfp2017 also doesn’t change dramatically, 549.fotonik3d does score quite a bit better than the M1, which could be tied to the more available DRAM bandwidth as this workloads puts extreme stress on the memory subsystem, but otherwise the scores change quite little compared to the M1, which is still on average quite ahead of the laptop competition.
The M1 Max lands as the top performing laptop chip in SPECint2017, just shy of being the best CPU overall which still goes to the 5950X, but is able to take and maintain the crown from the M1 in the FP suite.
Overall, the new M1 Max doesn’t deliver any large surprises on single-threaded performance metrics, which is also something we didn’t expect the chip to achieve.
Facebook
Twitter
RSS
493 Comments
View All Comments
Spunjji - Tuesday, October 26, 2021 - link
It's a combination of 2 of the responses you got - advanced packaging (having memory close to the SoC) and using the most advanced manufacturing process to spend a *lot* of transistors.On the CPU side, Zen 4 on TSMC N5 is likely to approach the power and performance levels of M1 Pro, albeit with a higher idle power and (probably) higher peak performance if efficiency is sacrificed entirely.
Ppietra - Tuesday, October 26, 2021 - link
more advanced node is a very small reason in order to explain it.AMD is supposedly already using the 7nm+ node and difference between this and 5nm is not that big... at most a 10% improvement in perf/W considering TSMC info.
As for the memory packaging advantage, as far as I understand that is an extra that is not included in these values (except in the Wall Power consumption), they are only measuring the processor package power. Even if you were talking about the memory controller the difference would only be 1-2W, hardly enough to explain the total difference in power consumption. And this is before actually looking at performance per Watt.
BushLin - Tuesday, October 26, 2021 - link
Combination microwave ovens are a thing btw, most efficient way to roast turkey if it fits inside. Just sayin'https://www.panasonic.com/uk/consumer/home-applian...
BushLin - Tuesday, October 26, 2021 - link
Was relevant to a comment above, now just looks weirdYouyou122 - Tuesday, October 26, 2021 - link
Is M1 max with 32 GB ram can reach to 400 GB bandwitch speed ? Or it’s just in 64 GB ram version?Spunjji - Tuesday, October 26, 2021 - link
AFAIK the 32GB version retains the full memory bandwidth.Spunjji - Tuesday, October 26, 2021 - link
It seems Apple's M1 chips are destined to produce flame-wars in the comments...I'm genuinely impressed by these designs. The CPU side of the SoC is extremely capable and, even accounting for node differences, shows a very interesting perf/watt over Intel and (to a lesser extent) AMD.
The GPU side, on the other hand, is less relevant to my interests. They've clearly designed an extremely capable arithmetic machine that's great for video filters. Unfortunately, between the near-non-existence of ARM binary games on Mac and the painful results of binary translation, it's apparently still not possible to game well on macOS. That perf/watt is still pretty promising, but they have less of an advantage over AMD/Nvidia here than on the CPU side. I suspect the transition to 5nm-class nodes will see AMD and Nvidia catching up again.
Farfolomew - Tuesday, October 26, 2021 - link
I was a bit disheartened as well with the M1's game performance. It kind of stinks that you have this incredible new SoC, that is 2-3x more efficient than anything before it, hamstrung by it's need to run binary translation. I suppose that was always going to be the case with Apple making the switch, but it's such a tease to see what a chip design could be truly capable given near infinite R&D $ and limitless design constraints.I'll be sticking to x86 for now. I hope the players in this camp step up their game and copy Apple in this new SoC design (or AMD produces a mainstream PC part that mimic's their console SoCs)
Tigran - Tuesday, October 26, 2021 - link
Why there are two different results of MSI GE76 Raider in GFXBench 5.0 Aztec Ruins High off-screen: 266 fps (page 'Power Behaviour') and 315 fps (page 'GPU Performance')?vladx - Tuesday, October 26, 2021 - link
Because Andrei is a terrible reviewer