Hi people need help. When i use vex in the solver to do caluclation.I found that AMD CPU is much slower than Intel CPU.
PC A spec: R9 5900X（PBO on），32Gb memory(3200mhz)，RTX3080.
PC B spec: i7 10700KF, 32Gb memory(2400mhz), RTX3070, All devices not overclocked.

When the size of grid is 200*200, 5900X work slowr than 10700KF. The 5900x has only two cores working, but all cores of 10700KF are working.
When the size of grid is 2000*2000, 5900X work faster than 10700KF.All cores of both CPUs are working.

BUT, when i disable the SMT(simultaneous multithreading) of AMD, no smt(5900x) is much faster than smt(200*200 points),but still slower than 10700KF.

Same thing on 5900x with your scene. Hope I was not mistaken in processor choice for houdini. Did you find a solution?

I'm getting around 20fps with 200x200pts
all cores are working on my 5950x
edit: on the h19.0.455

When we multi-thread VEX, we try to ensure there's enough work for all threads to be fully pegged. When there are only 4,000 points, we may not be able to saturate all the processors all the time. When there are 4,000,000 points, all the threads will kick in.

I see similar performance when I test on Intel vs. AMD locally -- though I'm running on Linux, and I don't have a 5950x.

I've spoken to AMD to see if they can reproduce this issue or shed more light on it.

JloIIIaDka
Same thing on 5900x with your scene. Hope I was not mistaken in processor choice for houdini. Did you find a solution?

yeah. I reinstall windwos system, and problem disappered.hahaha

mark
When we multi-thread VEX, we try to ensure there's enough work for all threads to be fully pegged. When there are only 4,000 points, we may not be able to saturate all the processors all the time. When there are 4,000,000 points, all the threads will kick in.

I see similar performance when I test on Intel vs. AMD locally -- though I'm running on Linux, and I don't have a 5950x.

I've spoken to AMD to see if they can reproduce this issue or shed more light on it.

thank a lot. i uodate windows to 11, and the problem seems to disappear.