kiroje wrote: ↑Fri Mar 04, 2022 9:25 am
after seeing the numbers for the 5800h ill go for that instead of the 5800u
I think that in order to make a good comparison it would be interesting to perform the same test. I used to calculate nps the command
bench 1024 16 16.
With this command the nps I think is somewhat lower, for example with 16 threads I think it is something like 16 Mnps.
On the page you can see that processors that have 8 cores/16 threads show significantly more nps when using 16 threads vs. 8 threads.
If there is something else that we may need to be taken into account, in this ranking they only take into account the nps, however in the bench command it also shows the time to complete the test. Perhaps when a processor with hyperthreading or smt exceeds the number of threads to cores for chess, the number of nodes per second increases but the time to complete the test does not decrease. So perhaps some people may be right that increasing the number of threads above 8 will not give more power to the engine. Here it would be interesting to see a match between stockfish 8 threads vs stockfish 16 trheads.
All this doesn't explain why the 5800U doesn't show more nps with 16 threads. The 5800U in turbo mode should have 8 cores at 3.40 GHz and the 5800H 8 cores at 3.80 GHz.
Frankly, from the posters stated rather minimal needs...I think just about anything would do. Forget NPS and such... and just search for a good deal on a mainstream laptop.
kiroje wrote: ↑Fri Mar 04, 2022 9:25 am
after seeing the numbers for the 5800h ill go for that instead of the 5800u
I think that in order to make a good comparison it would be interesting to perform the same test. I used to calculate nps the command
bench 1024 16 16.
With this command the nps I think is somewhat lower, for example with 16 threads I think it is something like 16 Mnps.
On the page you can see that processors that have 8 cores/16 threads show significantly more nps when using 16 threads vs. 8 threads.
If there is something else that we may need to be taken into account, in this ranking they only take into account the nps, however in the bench command it also shows the time to complete the test. Perhaps when a processor with hyperthreading or smt exceeds the number of threads to cores for chess, the number of nodes per second increases but the time to complete the test does not decrease. So perhaps some people may be right that increasing the number of threads above 8 will not give more power to the engine. Here it would be interesting to see a match between stockfish 8 threads vs stockfish 16 trheads.
All this doesn't explain why the 5800U doesn't show more nps with 16 threads. The 5800U in turbo mode should have 8 cores at 3.40 GHz and the 5800H 8 cores at 3.80 GHz.
ill go to the shop and try with that command, and return with some numbers.
“Modern chess is too much concerned with things like pawn structure. Forget it, checkmate ends the game.” – Nigel Short
pedrox wrote: ↑Fri Mar 04, 2022 12:18 pmAll this doesn't explain why the 5800U doesn't show more nps with 16 threads.
Cooling solution, as I said.
My HP software allows to configure 3 modes for the processor; silent mode, default and performance. Maybe the values given on the 5800U is because the silent mode (10W?) is selected. For 5800H:
bench 1024 8 26 default depth nnue
(silent mode)
Total time (ms) : 183496
Nodes searched : 1738509129
Nodes/second : 9474370
(default)
Total time (ms) : 128285
Nodes searched : 1572011822
Nodes/second : 12254057
bench 1024 16 26 default depth nnue
(silent mode)
Total time (ms) : 204061
Nodes searched : 2367240575
Nodes/second : 11600651
(default)
Total time (ms) : 164796
Nodes searched : 2593392687
Nodes/second : 15736988
Im looking for a new laptop that I will use primarily for chess. I have looked at an HP Envy with an AMD Ryzen 7 5800u 16Gb ram and it have a not to bad price.
But i tested it with the newest stockfisk and with 1 thread it gave around 1.8M nps - 4 threads around 4.2M nps and 8 threads it gave 8.6M nps
That is not too bad but when i tried with 12 and 16 threads it didnt give much more, 16 threads was around 9.3M nps - how can that be ?
Is it because of the size of the net and cache or something else (bandwidth problems etc.)
And which budget amd cpu is considered the best for chess ?
looking forward for answers.
Thanks
Buy the MacBook Pro 16-inch with M1 MAX chip.
Buy tomorrow the MacBook Pro with M2 chip.
pedrox wrote: ↑Fri Mar 04, 2022 12:18 pmAll this doesn't explain why the 5800U doesn't show more nps with 16 threads.
Cooling solution, as I said.
My HP software allows to configure 3 modes for the processor; silent mode, default and performance. Maybe the values given on the 5800U is because the silent mode (10W?) is selected. For 5800H:
bench 1024 8 26 default depth nnue
(silent mode)
Total time (ms) : 183496
Nodes searched : 1738509129
Nodes/second : 9474370
(default)
Total time (ms) : 128285
Nodes searched : 1572011822
Nodes/second : 12254057
bench 1024 16 26 default depth nnue
(silent mode)
Total time (ms) : 204061
Nodes searched : 2367240575
Nodes/second : 11600651
(default)
Total time (ms) : 164796
Nodes searched : 2593392687
Nodes/second : 15736988
I now tested my AMD Ryzen 5800HS and get these numbers with the extreme performance settings
Thank you for your test. The 5800hs is like the 5800h with a slightly lower base frequency and a TDP of 35W vs. 45W for the other. Values appear to be in line with expectations.
I would like to see values of some AMD u-series; especially the 4800u, 5600u, 5700u and 5800u again by another user to see if it behaves the same as the presented values.