Best cpu benchmark for chess?

[h1a8]

I propose we do a study in order to find best benchmark (geekbench 5 integer, cinebench, etc) that highly correlates to chess kn/s (stockfish 14 as the standard).

I have 3 computers
A desktop running i7-9700k
A laptop running i7-11800h i just purchased
Another laptop I just purchased for my daughter (i5-1135g7)

I had a 10th gen i7 laptop but gave it away.

I did a regression analysis for the 3 computers (bench vs geekbench integer). I got r^2 = 0.9675

The regression equation is y = 3.69x -8346 where x is geekbench integer score and y is the bench (command line "bench 1024 xx 26" where xx is the number of threads).

Here are my numbers
i7-9700k 16509kn/s bench and 6317kn/s geekbench integer
i5-1135g7 6934kn/s bench and 4351kn/s geekbench integer
i7-11800h 21538kn/s bench and 8306kn/s geekbench integer

I haven't did a regression on cinebench yet.

Feel free to add your numbers here to get a more accurate regression analysis.

[Sopel]

https://openbenchmarking.org/test/pts/stockfish
http://ipmanchess.yolasite.com/amd--int ... ckfish.php
why do other benchmarks when you want to know chess nps?

[h1a8]

https://openbenchmarking.org/test/pts/stockfish
http://ipmanchess.yolasite.com/amd--int ... ckfish.php
why do other benchmarks when you want to know chess nps?

The goal was to determine chess strength from m1 max, android, ios, etc (non pc) with geekbench integer alone (or another universal benchmark program).

[yurikvelo]

Why you test ratio on x86-64/AVX2 CPU if you want ARM CPU ratio

[Milos]

https://openbenchmarking.org/test/pts/stockfish
http://ipmanchess.yolasite.com/amd--int ... ckfish.php
why do other benchmarks when you want to know chess nps?

The goal was to determine chess strength from m1 max, android, ios, etc (non pc) with geekbench integer alone (or another universal benchmark program).

Oh for god sake, those marketing synthetic benchmarks (particularly geekbench) are terrible prediction of performance for chess. Particularly in ARM based Apple architectures (Apple and most cellphone manufacturers simply insert shortcuts in their hardware just to perform on particular benchmarks and then optimize compiles particularly for them).
None of them measures extended set of instructions, SIMD, branching prediction performance and integer type of instructions used in chess (bitboards manipulation ones, i.e. Geekbench integer is useless there).
So what you are looking for is simply impossible with those benchmarks.

[JohnWoe]

geekbench is boring as f
SF bench is propably the best for chess benching.

I developed a Collatz conjecture bench program. Numbers are bitboards/u64 so quite fitting to chess. Utilizes all CPUs you have
https://github.com/SamuraiDangyo/collatzzz
Hash can be set too. But steps are pretty small on Collatz so it won't help much. Or I haven't found a good way to get boos from hash. But measures RAM speed too.
Here v0.3 (Ryzen 4800U):

Code: Select all

./collatzzz -sum 0 1000000000

Code: Select all

Collatz:    0 -> 1,000,000,000
Sum(steps): 203,234,783,374
CPU(s):     16
Hash(MB):   0
SPS:        8,338,521,453
Time(ms):   24,373

[towforce]

If there's a strong correlation between Geekbench and chess benchmarking programs (I suspect there will be), it will be extremely useful because there is public information on the Geekbench scores for nearly all the CPUs and SOCs available.

The effort here to confirm or disprove the strong correlation is worthy of praise!

[Milos]

If there's a strong correlation between Geekbench and chess benchmarking programs (I suspect there will be), it will be extremely useful because there is public information on the Geekbench scores for nearly all the CPUs and SOCs available.

The effort here to confirm or disprove the strong correlation is worthy of praise!

How many times one has to repeat. There is NO correlation between Geekbench (integer or overall) results and chess program benchmarks.

[towforce]

If there's a strong correlation between Geekbench and chess benchmarking programs (I suspect there will be), it will be extremely useful because there is public information on the Geekbench scores for nearly all the CPUs and SOCs available.

The effort here to confirm or disprove the strong correlation is worthy of praise!

How many times one has to repeat. There is NO correlation between Geekbench (integer or overall) results and chess program benchmarks.

If someone has already done the study which the OP is proposing, I was not aware of it.

[Milos]

If there's a strong correlation between Geekbench and chess benchmarking programs (I suspect there will be), it will be extremely useful because there is public information on the Geekbench scores for nearly all the CPUs and SOCs available.

The effort here to confirm or disprove the strong correlation is worthy of praise!

How many times one has to repeat. There is NO correlation between Geekbench (integer or overall) results and chess program benchmarks.

If someone has already done the study which the OP is proposing, I was not aware of it.

A study, seriously???
There are 3 points in this thread, there are a couple of points for M1 CPU, anyone can generate numbers for their own CPU they have. So your "study" is to collect those data do a linear regression and get a correlation coefficient.
Ofc doing it when data points form a swarm is a total waste of time, but you are free to make that "study".

P.S. OP did a linear regression on 3 data points and made some conclusions. That should be enough even for a scientifically aware high school kid to realize what kind of pointless and clueless discussion this whole thread is.