x3d cache vs x cpu

[landjalan]

hi guys, do you think that an 5800x3d vs 5800x would be faster in calculating prime numbers ?

i did not find benchmarks of any x3d chip to compare.

thanks for any input

[kruoli]

5800X and 5800X3D.

[moebius]

Originally Posted by kruoli

5800X and 5800X3D.

The 5800X scales slightly better according to the benchmarks.

[sdbardwick]

FFT Benchmark for 5800X3D:

AMD Ryzen 7 5800X3D 8-Core Processor           
CPU speed: 3400.15 MHz, 8 hyperthreaded cores
CPU features: 3DNow! Prefetch, SSE, SSE2, SSE4, AVX, AVX2, FMA
L1 cache size: 8x32 KB, L2 cache size: 8x512 KB, L3 cache size: 96 MB
L1 cache line size: 64 bytes, L2 cache line size: 64 bytes
Machine topology as determined by hwloc library:
 Machine#0 (total=13812504KB, Backend=Windows, OSName=Windows, WindowsBuildEnvironment=MinGW, OSRelease=10, OSVersion=10.0.19045, Hostname=5800X3D, Architecture=x86_64, hwlocVersion=2.4.1, ProcessName=prime95.exe)
  Package (total=13812504KB, CPUVendor=AuthenticAMD, CPUFamilyNumber=25, CPUModelNumber=33, CPUModel="AMD Ryzen 7 5800X3D 8-Core Processor           ", CPUStepping=2)
    L3 (size=98304KB, linesize=64, ways=16, Inclusive=0)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x00000003)
            PU#0 (cpuset: 0x00000001)
            PU#1 (cpuset: 0x00000002)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x0000000c)
            PU#2 (cpuset: 0x00000004)
            PU#3 (cpuset: 0x00000008)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x00000030)
            PU#4 (cpuset: 0x00000010)
            PU#5 (cpuset: 0x00000020)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x000000c0)
            PU#6 (cpuset: 0x00000040)
            PU#7 (cpuset: 0x00000080)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x00000300)
            PU#8 (cpuset: 0x00000100)
            PU#9 (cpuset: 0x00000200)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x00000c00)
            PU#10 (cpuset: 0x00000400)
            PU#11 (cpuset: 0x00000800)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x00003000)
            PU#12 (cpuset: 0x00001000)
            PU#13 (cpuset: 0x00002000)
      L2 (size=512KB, linesize=64, ways=8, Inclusive=1)
        L1d (size=32KB, linesize=64, ways=8, Inclusive=0)
          Core (cpuset: 0x0000c000)
            PU#14 (cpuset: 0x00004000)
            PU#15 (cpuset: 0x00008000)
Prime95 64-bit version 30.7, RdtscTiming=1
Timing FFTs using 8 cores.
Best time for 1024K FFT length: 0.536 ms., avg: 0.543 ms.
Best time for 1120K FFT length: 0.593 ms., avg: 0.598 ms.
Best time for 1152K FFT length: 0.583 ms., avg: 0.588 ms.
Best time for 1280K FFT length: 0.669 ms., avg: 0.677 ms.
Best time for 1344K FFT length: 0.700 ms., avg: 0.706 ms.
Best time for 1440K FFT length: 0.736 ms., avg: 0.741 ms.
Best time for 1536K FFT length: 0.802 ms., avg: 0.812 ms.
Best time for 1600K FFT length: 0.809 ms., avg: 0.815 ms.
Best time for 1680K FFT length: 0.891 ms., avg: 0.896 ms.
Best time for 1792K FFT length: 0.959 ms., avg: 0.965 ms.
Best time for 1920K FFT length: 0.965 ms., avg: 0.971 ms.
Best time for 2048K FFT length: 1.102 ms., avg: 1.112 ms.
Best time for 2240K FFT length: 1.184 ms., avg: 1.189 ms.
Best time for 2304K FFT length: 1.170 ms., avg: 1.175 ms.
Best time for 2400K FFT length: 1.253 ms., avg: 1.256 ms.
Best time for 2560K FFT length: 1.393 ms., avg: 1.401 ms.
Best time for 2688K FFT length: 1.422 ms., avg: 1.430 ms.
Best time for 2800K FFT length: 1.524 ms., avg: 1.530 ms.
Best time for 2880K FFT length: 1.491 ms., avg: 1.497 ms.
Best time for 3072K FFT length: 1.575 ms., avg: 1.581 ms.
Best time for 3200K FFT length: 1.689 ms., avg: 1.699 ms.
Best time for 3360K FFT length: 1.808 ms., avg: 1.813 ms.
Best time for 3584K FFT length: 1.908 ms., avg: 1.929 ms.
Best time for 3840K FFT length: 2.056 ms., avg: 2.066 ms.
Best time for 4096K FFT length: 2.211 ms., avg: 2.221 ms.
Best time for 4480K FFT length: 2.456 ms., avg: 2.465 ms.
Best time for 4608K FFT length: 2.433 ms., avg: 2.446 ms.
Best time for 4800K FFT length: 2.620 ms., avg: 2.630 ms.
Best time for 5120K FFT length: 2.774 ms., avg: 2.783 ms.
Best time for 5376K FFT length: 3.007 ms., avg: 3.027 ms.
Best time for 5600K FFT length: 3.012 ms., avg: 3.019 ms.
Best time for 5760K FFT length: 3.048 ms., avg: 3.058 ms.
Best time for 6144K FFT length: 3.353 ms., avg: 3.364 ms.
Best time for 6400K FFT length: 3.427 ms., avg: 3.438 ms.
Best time for 6720K FFT length: 3.592 ms., avg: 3.605 ms.
Best time for 7168K FFT length: 4.048 ms., avg: 4.056 ms.
Best time for 7680K FFT length: 4.260 ms., avg: 4.283 ms.
Best time for 8000K FFT length: 4.364 ms., avg: 4.375 ms.
Best time for 8064K FFT length: 4.528 ms., avg: 4.539 ms.
Best time for 8192K FFT length: 4.655 ms., avg: 4.665 ms.
Best time for 8960K FFT length: 5.172 ms., avg: 5.195 ms.
Best time for 9216K FFT length: 5.059 ms., avg: 5.078 ms.
Best time for 9600K FFT length: 5.507 ms., avg: 5.614 ms.
Best time for 10240K FFT length: 6.021 ms., avg: 6.060 ms.
Best time for 10752K FFT length: 6.014 ms., avg: 6.033 ms.
Best time for 11200K FFT length: 6.369 ms., avg: 6.449 ms.
Best time for 11520K FFT length: 7.175 ms., avg: 7.435 ms.
Best time for 12288K FFT length: 6.893 ms., avg: 6.974 ms.
Best time for 12800K FFT length: 8.641 ms., avg: 8.837 ms.
Best time for 13440K FFT length: 10.045 ms., avg: 10.120 ms.
Best time for 13824K FFT length: 10.283 ms., avg: 10.426 ms.
Best time for 14336K FFT length: 11.526 ms., avg: 11.708 ms.
Best time for 15360K FFT length: 12.678 ms., avg: 12.803 ms.
Best time for 16000K FFT length: 10.767 ms., avg: 10.839 ms.
Best time for 16128K FFT length: 14.446 ms., avg: 14.557 ms.
Best time for 16384K FFT length: 15.579 ms., avg: 15.650 ms.
Best time for 17920K FFT length: 12.663 ms., avg: 12.725 ms.
Best time for 18432K FFT length: 14.541 ms., avg: 14.633 ms.
Best time for 19200K FFT length: 14.682 ms., avg: 14.797 ms.
Best time for 20480K FFT length: 16.808 ms., avg: 16.931 ms.
Best time for 21504K FFT length: 20.451 ms., avg: 20.711 ms.
Best time for 22400K FFT length: 20.165 ms., avg: 20.373 ms.
Best time for 23040K FFT length: 22.950 ms., avg: 23.120 ms.

[xx005fs]

Originally Posted by sdbardwick

FFT Benchmark for 5800X3D:

Solid results for the 5800X3D, but I think the real speed up would be with the 7800X3D (or other zen4 X3D chips) with AVX512 and no bandwidth constraint. I have a 7950X3D run in the perpetual benchmark thread, and also getting between 1.6-1.8ms/it on the V-Cache CCD alone for 6144K (wavefront) FFT.

[Mark Rose]

Originally Posted by landjalan

hi guys, do you think that an 5800x3d vs 5800x would be faster in calculating prime numbers ?

i did not find benchmarks of any x3d chip to compare.

thanks for any input

Not worth paying for an upgrade for prime hunting.

Definitely worth it for many games.