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
Ppietra - Monday, October 25, 2021 - link
The number of transistors doesn’t say what is price because, like you said, they are not manufactured on the same node, nor even by the same company, so you have no idea what is the difference in yields for the manufacturing process used.Secondly the 3090 is actually much bigger than the M1 Max (630 mm^2 vs estimated 430 mm^2), and it’s the area that affects the yield per wafer on that node.
Thirdly you still have no idea what is the actual cost of the 3090 chip, so you don’t have a clue for the price of M1 Max to base your argument on.
sirmo - Monday, October 25, 2021 - link
TSMC discloses scaling between different nodes (84% better density from 7nm-5nm). So we have a pretty good idea. I am 100% sure M1 Max is more expensive to manufacture than 3090.Ppietra - Monday, October 25, 2021 - link
:S what are you talking about!??? We know the size of the 3090 and we already have an estimate for the M1 Max based on the size of M1. The 3090 is much much bigger.And it is not enough to be more expensive than the 3090, you actually need a very high cost value for your argument about subsidies to make any sense - and by the way you haven’t given any value that would make a case for it!
sirmo - Monday, October 25, 2021 - link
5nm is 84% denser than 7nm, and even more so compared to Samsung's 8nm. Which is why transistor count is easier to understand.Ppietra - Monday, October 25, 2021 - link
you continue to make no sense! What matters when talking about yields is die size. Transistors count doesn’t tell you anything about yields when comparing chips using different manufacturing processes.vlad42 - Monday, October 25, 2021 - link
The price per transistor has generally been on the rise since post 14nm. So, just because the 3090 is larger, that does not mean it is more expensive.Also, I would not say the M1 Max MacBook Pro is being sold at a loss, but I would bet that the upgrade cost from the M1 Pro is less than then cost difference between the two chips. That is to say, I think Apple's margins on the M1 Max MacBook Pro are less than the M1 Pro MacBook Pro (or the margins from the other extra components such as memory are making up for it).
valuearb - Tuesday, October 26, 2021 - link
The higher the price point the higher Apples margins are, that’s always been true. The margin on the $1999 14 inch M1 Pro MBP is always going to be less than margins on a $3500 16 inch M1 Max MBP.Think about it. It’s basic pricing strategy. The high end buyers aren’t as price sensitive, have higher disposable income, earn more from their MBP investment, etc.
Ppietra - Tuesday, October 26, 2021 - link
vlad42, never have I said that the 3090 is more expensive... my argument is that you cannot say that the M1 Max is more expensive just because it has more transistors - yields are affected by dye area which affects the price and there are other factors that affect price too. As for the price per transistor info, that is assuming the yield is always the same.They charge 400 dollars more just to upgrade to the M1 Max. Considering the size of the M1 Max that would give around 130 chips per wafer, the price per wafer should be around 17,000 dollars (according to some estimates). That would mean 130 dollars per chip at 100% yield. An extra 400 dollars should be more than enough to cover a low yields and packaging.
valuearb - Tuesday, October 26, 2021 - link
And Microsoft puts 15B transistors in a $500 gaming device. I’m thinking a $3,000 laptop has plenty of margin to cover those big wafers.valuearb - Tuesday, October 26, 2021 - link
Apple never sells at a loss, always prices for high margins, and has zero reason to change that now given the new MBPs were instantly back ordered. Underpricing would be leaving mass amounts of money on the table.