Jump to content

Can m1 beat core i9 10900k in single core performance?

I was curious to compare new M1 processor with AMD and Intel processors I watch some videos and read some articles on web. In one article I find out that M1 can beat core i9 10900k in single core performance. I don't think it real but can any verify it. 

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, Ibrahim habib said:

I was curious to compare new M1 processor with AMD and Intel processors I watch some videos and read some articles on web. In one article I find out that M1 can beat core i9 10900k in single core performance. I don't think it real but can any verify it. 

That's geekbench, and geekbench isn't a "great" benchmark. It's more optimized for macs, feel free to correct me if im wrong

geometry is hard
b550 > x570

Link to comment
Share on other sites

Link to post
Share on other sites

It would be very dependent on what you are doing, there are probably some M1 wins but I would expect the i9 to be better across a wider spectrum

Link to comment
Share on other sites

Link to post
Share on other sites

21 minutes ago, Downkey said:

That's geekbench, and geekbench isn't a "great" benchmark. It's more optimized for macs, feel free to correct me if im wrong

The link he gave also has cinebench scores, which the M1 beats in single core too

Link to comment
Share on other sites

Link to post
Share on other sites

i think both are good for their own use. I dont think gaming will be great on an arm processor any time soon. Then again I think they are parts of your gpus and even some part of the cpu. Nonetheless, this will be good for consumers. Imagine 4 core 8 threads in 2017 was the highest consumer cpu available (other then the x299 and high prices)

Link to comment
Share on other sites

Link to post
Share on other sites

M1 is technically better, yes

please quote me or tag me @wall03 so i can see your response

motherboard buying guide      psu buying guide      pc building guide     privacy guide

ltt meme thread

folding at home stats

 

pc:

 

RAM: 16GB DDR4-3200 CL-16

CPU: AMD Ryzen 5 3600 @ 3.6GHz

SSD: 256GB SP

GPU: Radeon RX 570 8GB OC

OS: Windows 10

Status: Main PC

Cinebench R23 score: 9097 (multi) 1236 (single)

 

don't some things look better when they are lowercase?

-wall03

 

hello dark mode users

goodbye light mode users

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, wall03 said:

M1 is technically better, yes

Wins however are not made on technicalities, the 10900K floor stomps the M1 into Morrowind and back through the gates of Oblivion.

My computer for gaming & work. AMD Ryzen 3600x with XFR support on - Arctic Cooling LF II - ASUS Prime X570-P - Gigabyte 5700XT - 32GB Geil Orion 3600 - Crucial P1 1TB NVME - Crucial BX 500 SSD - EVGA GQ 650w - NZXT Phantom 820 Gun Metal Grey colour - Samsung C27FG73FU monitor - Blue snowball mic - External best connectivity 24 bit/ 96khz DAC headphone amp -Pioneer SE-205 headphone - Focal Auditor 130mm speakers in custom sealed boxes - inPhase audio XT 8 V2 wired at 2ohm 300RMS custom slot port compact box - Vibe Audio PowerBox 400.1

Link to comment
Share on other sites

Link to post
Share on other sites

2 hours ago, Ibrahim habib said:

I was curious to compare new M1 processor with AMD and Intel processors I watch some videos and read some articles on web. In one article I find out that M1 can beat core i9 10900k in single core performance. I don't think it real but can any verify it. 

 

Geekbench is fine, but you have to understand the shortcomings of synthetic benchmarks

a) They generate binaries for different CPU types, which ARM is Big Endian and Intel is Little Endian, so there are byte-swapping instructions needed when translating between the two, that native doesn't need, but the data used for the benchmark may require to not ship two separate data sets.

b) They generate binaries for different operating systems, which may have different threading schedulers

c) Things like RAM and GPU are not universally comparable.

 

When you compare two CPU's, for CPU performance alone, you're omitting the other parts of the SoC in the ARM chip, so in effect you're penalizing the SoC but not the dedicated CPU.

 

The closest thing the M1 can be compared to is the Intel CPU with it's onboard GPU, and AMD's G parts (eg Ryzen 7 4700G, which these parts are based on the previous generation of CPU core.) So you can not compare them to Xeon's or other Intel parts that do not have GPU's. You have to do a comparison of like-for-like, and that would require using only the OpenCL performance on the CPU + GPU and not the SIMD instructions (eg AVX on x86-64 parts.) That means it will run into driver optimization issues.

 

So while it's easy to compare two CPU's of the same type by eliminating all the different variables, the only way to get a fair comparison is to compare the chips with the same software on the same operating system on the same hardware, which doesn't really exist, and even going the hackintosh route to do so doesn't give you a fair comparison because you can't compare AMD's chips then.

 

Cinebench, likewise is fair to the same extent geekbench is. 

 

I would not make a purchasing decision based on synthetic benchmarks. They only serve to rank CPU's of the same architecture on the same platform, and while it is true the M1 laps some Intel scores, that's not true for every single thing, particularly multicore. When you start comparing things with un-equal CPU cores, not even the thermal conditions are the same.

 

Hence it's only fair to compare the M1 to the ultrabook CPU's, and that is where the M1 laps the Intel parts significantly.

 

https://www.extremetech.com/computing/317304-benchmark-results-show-apple-m1-beating-every-intel-powered-macbook-pro

Quote

The highest-end MacBook Pro 16-inch that Apple has ever sold features a Core i9-9980HK, an 8C/16T CPU with a base clock of 2.4GHz and a 5GHz boost. Measured performance for this CPU is 1149 / 7329 in GB5. Now, this is an older 14nm CPU. Intel doesn’t currently ship an eight-core Ice Lake mobile CPU, but Apple also ships Intel’s quad-core, ICL-derived Core i7-1068NG7. That chip scores 1352 / 4914. Apple is therefore leading in single-threaded performance even against Intel’s Sunny Cove architecture from 2019, and it’s leading in multi-threaded against the highest performing CPU cores that Intel ships in mobile. While Apple is only tying the Core i9-9980HK in multi-threaded, the M1’s Geekbench scores cleanly beat everything below that point, including the Core i9-9880H.

 

When you start comparing it to the desktop CPU's, the sustained performance will not be fair, as mobile parts are designed to stick to a thermal envelope (in this case 15w) where a desktop might pull 20x as much under the same benchmark. Especially with a desktop GPU.

 

Link to comment
Share on other sites

Link to post
Share on other sites

Yes, the M1 has more powerful cores than the 10900K. There is more to a chip than just single core performance though. 

The Intel processor has far more cores so if you are doing workloads that can scale beyond 4 cores the Intel processor will most likely pull ahead. The M1 however uses less power and has a more powerful GPU, as well as a dedicated NPU. The Intel chip pulls ahead when we look at connectivity (more specifically the number of PCIe lanes). 

 

If we only focus on single core performance and ignores everything else, then the M1 is better. 

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×