Perpetual benchmark thread... - Page 74

kladner · 2020-05-02, 06:09

Quote:

Originally Posted by Mark Rose

Your 1 worker results are far faster than your 2 worker results. I'd run only a single worker after seeing your results.

The higher core count Ryzen and Threadripper parts are very memory bandwidth starved for Prime95 (same with Intel parts). You might get a little more performance with dual rank memory, or four sticks of memory.

I'd be curious to see how 1 worker with 6, 7, and 8 cores performs. The sweet spot for the available memory bandwidth is probably around 7 or 8 cores.

Dual rank, or 4 DIMMS in a dual channel system can boost almost 30%. I might be able to test this by shutting off one pair of single rank on a system here. That machine was painfully slower than expected out of the box, ~~though I would have to dig for benchmarks~~ there are before and after adding 2 more DIMMs benchmarks on that machine. I'll get those and post, though I was only doing Throughput benches. I think they are mostly 'try all combinations' runs.

Viliam Furik · 2020-05-02, 22:26

I've recently run a few throughput benchmarks on the 3900X.
I attached graphs of the results.

The first one is normalized throughput benchmark from size 64K to 9600K (only on a single worker, other worker settings are up to 6400K) for 1, 2 4, and 12 workers, all using 12 cores in total. Normalization was done by the following formula [normperf = throughput * ln(FFT size) * FFT size / clock speed (4,1 GHz)]. You can see performance drops, which show when the cache runs out. It's possible that values correspond to the number of double-precision floating-point operations per cycle, which for Zen 2 is 16, under ideal conditions.

The second graph is a single worker throughput using 1, 2, 3, ..., 12 cores, showed in iterations per second. I suspect the "bump" about 6 cores has something to do with 3900X having two 6-core chiplets. Perhaps when using cores from two chiplets, it has access to cache from both of them.

ZIP file contains the values used, as well as original graphs, in an Excel spreadsheet.

Mark Rose · 2020-05-03, 04:11

Quote:

Originally Posted by Viliam Furik

I've recently run a few throughput benchmarks on the 3900X.
I attached graphs of the results.

Very insightful! What's your memory configuration?

LaurV · 2020-05-03, 05:21

Quote:

Originally Posted by Viliam Furik

I've recently run ...

Nicely done and quite helpful.

Viliam Furik · 2020-05-03, 12:38

Quote:

Originally Posted by Mark Rose

Very insightful! What's your memory configuration?

4 * 16 GB 3200 MHz

moebius · 2020-05-04, 14:04

Throughput-Test:

Code:

AMD Ryzen 7 3700X 8-Core Processor             
CPU speed: 4299.72 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: 2x16 MB
L1 cache line size: 64 bytes, L2 cache line size: 64 bytes
Machine topology as determined by hwloc library:
 Machine#0 (total=14412868KB, Backend=Windows, hwlocVersion=2.0.4, ProcessName=prime95.exe)
  Package (total=14412868KB, CPUVendor=AuthenticAMD, CPUFamilyNumber=23, CPUModelNumber=113, CPUModel="AMD Ryzen 7 3700X 8-Core Processor             ", CPUStepping=0)
    L3 (size=16384KB, 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)
    L3 (size=16384KB, 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: 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 29.8, RdtscTiming=1
Timings for 2048K FFT length (8 cores, 1 worker):  1.24 ms.  Throughput: 807.32 iter/sec.
Timings for 2048K FFT length (8 cores, 2 workers):  2.20,  2.14 ms.  Throughput: 920.77 iter/sec.
Timings for 2048K FFT length (8 cores, 8 workers): 24.56, 24.02, 23.76, 23.82, 23.76, 23.75, 23.80, 23.92 ms.  Throughput: 334.42 iter/sec.
Timings for 2048K FFT length (8 cores hyperthreaded, 1 worker):  1.26 ms.  Throughput: 795.21 iter/sec.
Timings for 2048K FFT length (8 cores hyperthreaded, 2 workers):  2.50,  2.43 ms.  Throughput: 812.25 iter/sec.
Timings for 2048K FFT length (8 cores hyperthreaded, 8 workers): 26.18, 24.42, 24.43, 24.18, 25.13, 25.00, 24.92, 24.36 ms.  Throughput: 322.42 iter/sec.
Timings for 2240K FFT length (8 cores, 1 worker):  1.33 ms.  Throughput: 750.82 iter/sec.
Timings for 2240K FFT length (8 cores, 2 workers):  3.09,  3.09 ms.  Throughput: 646.34 iter/sec.
Timings for 2240K FFT length (8 cores, 8 workers): 26.39, 26.55, 26.28, 26.14, 26.30, 26.07, 26.30, 26.48 ms.  Throughput: 304.01 iter/sec.
Timings for 2240K FFT length (8 cores hyperthreaded, 1 worker):  1.23 ms.  Throughput: 813.08 iter/sec.
Timings for 2240K FFT length (8 cores hyperthreaded, 2 workers):  3.42,  3.29 ms.  Throughput: 596.25 iter/sec.
Timings for 2240K FFT length (8 cores hyperthreaded, 8 workers): 28.39, 27.13, 27.01, 27.37, 28.03, 27.03, 27.53, 27.03 ms.  Throughput: 291.64 iter/sec.
Timings for 2304K FFT length (8 cores, 1 worker):  1.36 ms.  Throughput: 733.64 iter/sec.
Timings for 2304K FFT length (8 cores, 2 workers):  3.46,  3.38 ms.  Throughput: 584.60 iter/sec.
Timings for 2304K FFT length (8 cores, 8 workers): 27.34, 27.50, 27.20, 26.93, 27.31, 27.15, 26.95, 27.31 ms.  Throughput: 294.01 iter/sec.
Timings for 2304K FFT length (8 cores hyperthreaded, 1 worker):  1.30 ms.  Throughput: 768.02 iter/sec.
Timings for 2304K FFT length (8 cores hyperthreaded, 2 workers):  3.69,  3.57 ms.  Throughput: 551.00 iter/sec.
Timings for 2304K FFT length (8 cores hyperthreaded, 8 workers): 29.36, 27.92, 27.77, 27.85, 28.54, 27.87, 27.88, 27.98 ms.  Throughput: 284.31 iter/sec.
Timings for 2400K FFT length (8 cores, 1 worker):  1.46 ms.  Throughput: 684.85 iter/sec.
[Mon May 04 14:31:35 2020]
Timings for 2400K FFT length (8 cores, 2 workers):  4.10,  4.03 ms.  Throughput: 491.95 iter/sec.
Timings for 2400K FFT length (8 cores, 8 workers): 28.47, 28.47, 28.58, 28.81, 28.75, 28.40, 28.81, 28.40 ms.  Throughput: 279.87 iter/sec.
Timings for 2400K FFT length (8 cores hyperthreaded, 1 worker):  1.32 ms.  Throughput: 757.62 iter/sec.
Timings for 2400K FFT length (8 cores hyperthreaded, 2 workers):  4.37,  4.25 ms.  Throughput: 464.38 iter/sec.
Timings for 2400K FFT length (8 cores hyperthreaded, 8 workers): 30.96, 29.13, 29.47, 29.33, 30.25, 29.19, 29.54, 29.59 ms.  Throughput: 269.62 iter/sec.
Timings for 2560K FFT length (8 cores, 1 worker):  1.54 ms.  Throughput: 650.82 iter/sec.
Timings for 2560K FFT length (8 cores, 2 workers):  4.55,  4.51 ms.  Throughput: 441.91 iter/sec.
Timings for 2560K FFT length (8 cores, 8 workers): 29.91, 30.15, 29.88, 29.96, 29.85, 29.88, 29.93, 29.97 ms.  Throughput: 267.20 iter/sec.
Timings for 2560K FFT length (8 cores hyperthreaded, 1 worker):  1.41 ms.  Throughput: 709.08 iter/sec.
Timings for 2560K FFT length (8 cores hyperthreaded, 2 workers):  5.01,  4.87 ms.  Throughput: 404.83 iter/sec.
Timings for 2560K FFT length (8 cores hyperthreaded, 8 workers): 32.29, 31.39, 30.68, 30.56, 31.75, 30.48, 30.86, 30.79 ms.  Throughput: 257.33 iter/sec.
Timings for 2688K FFT length (8 cores, 1 worker):  1.65 ms.  Throughput: 605.84 iter/sec.
Timings for 2688K FFT length (8 cores, 2 workers):  5.37,  5.29 ms.  Throughput: 375.15 iter/sec.
Timings for 2688K FFT length (8 cores, 8 workers): 31.93, 31.95, 31.93, 31.61, 31.89, 32.03, 31.87, 31.83 ms.  Throughput: 250.93 iter/sec.
Timings for 2688K FFT length (8 cores hyperthreaded, 1 worker):  1.63 ms.  Throughput: 613.26 iter/sec.
Timings for 2688K FFT length (8 cores hyperthreaded, 2 workers):  5.73,  5.61 ms.  Throughput: 352.64 iter/sec.
Timings for 2688K FFT length (8 cores hyperthreaded, 8 workers): 34.20, 32.32, 32.46, 32.63, 33.43, 32.30, 33.15, 32.54 ms.  Throughput: 243.42 iter/sec.
Timings for 2800K FFT length (8 cores, 1 worker):  1.68 ms.  Throughput: 594.12 iter/sec.
Timings for 2800K FFT length (8 cores, 2 workers):  6.15,  6.08 ms.  Throughput: 327.19 iter/sec.
Timings for 2800K FFT length (8 cores, 8 workers): 33.44, 33.98, 33.69, 33.61, 33.70, 33.63, 33.69, 33.61 ms.  Throughput: 237.61 iter/sec.
[Mon May 04 14:36:45 2020]
Timings for 2800K FFT length (8 cores hyperthreaded, 1 worker):  1.60 ms.  Throughput: 626.65 iter/sec.
Timings for 2800K FFT length (8 cores hyperthreaded, 2 workers):  6.31,  6.16 ms.  Throughput: 320.72 iter/sec.
Timings for 2800K FFT length (8 cores hyperthreaded, 8 workers): 38.11, 35.42, 34.76, 33.58, 36.76, 34.52, 34.56, 34.19 ms.  Throughput: 227.37 iter/sec.
Timings for 2880K FFT length (8 cores, 1 worker):  1.73 ms.  Throughput: 579.30 iter/sec.
Timings for 2880K FFT length (8 cores, 2 workers):  6.25,  6.20 ms.  Throughput: 321.18 iter/sec.
Timings for 2880K FFT length (8 cores, 8 workers): 34.57, 34.26, 34.36, 34.59, 34.58, 34.02, 34.55, 34.65 ms.  Throughput: 232.24 iter/sec.
Timings for 2880K FFT length (8 cores hyperthreaded, 1 worker):  1.61 ms.  Throughput: 622.06 iter/sec.
Timings for 2880K FFT length (8 cores hyperthreaded, 2 workers):  6.85,  6.77 ms.  Throughput: 293.71 iter/sec.
Timings for 2880K FFT length (8 cores hyperthreaded, 8 workers): 36.99, 35.06, 35.40, 35.21, 36.59, 35.29, 35.79, 35.75 ms.  Throughput: 223.78 iter/sec.
Timings for 3072K FFT length (8 cores, 1 worker):  1.85 ms.  Throughput: 540.14 iter/sec.
Timings for 3072K FFT length (8 cores, 2 workers):  7.00,  7.00 ms.  Throughput: 285.84 iter/sec.
Timings for 3072K FFT length (8 cores, 8 workers): 35.93, 36.25, 35.92, 36.08, 35.82, 35.94, 36.00, 35.83 ms.  Throughput: 222.40 iter/sec.
Timings for 3072K FFT length (8 cores hyperthreaded, 1 worker):  1.87 ms.  Throughput: 534.95 iter/sec.
Timings for 3072K FFT length (8 cores hyperthreaded, 2 workers):  7.48,  7.34 ms.  Throughput: 269.93 iter/sec.
Timings for 3072K FFT length (8 cores hyperthreaded, 8 workers): 39.03, 37.33, 37.36, 37.31, 38.33, 37.27, 37.37, 37.24 ms.  Throughput: 212.50 iter/sec.
Timings for 3200K FFT length (8 cores, 1 worker):  2.06 ms.  Throughput: 485.25 iter/sec.
Timings for 3200K FFT length (8 cores, 2 workers):  7.52,  7.47 ms.  Throughput: 266.74 iter/sec.
Timings for 3200K FFT length (8 cores, 8 workers): 38.07, 38.28, 37.78, 38.05, 38.11, 38.09, 38.06, 38.06 ms.  Throughput: 210.18 iter/sec.
Timings for 3200K FFT length (8 cores hyperthreaded, 1 worker):  2.00 ms.  Throughput: 499.75 iter/sec.
Timings for 3200K FFT length (8 cores hyperthreaded, 2 workers):  7.84,  7.80 ms.  Throughput: 255.82 iter/sec.
[Mon May 04 14:41:58 2020]
Timings for 3200K FFT length (8 cores hyperthreaded, 8 workers): 41.18, 39.00, 39.10, 38.65, 40.45, 38.64, 38.89, 38.75 ms.  Throughput: 203.50 iter/sec.
Timings for 3360K FFT length (8 cores, 1 worker):  2.18 ms.  Throughput: 457.83 iter/sec.
Timings for 3360K FFT length (8 cores, 2 workers):  8.19,  8.16 ms.  Throughput: 244.54 iter/sec.
Timings for 3360K FFT length (8 cores, 8 workers): 40.60, 40.41, 39.87, 40.59, 39.98, 40.35, 40.48, 40.45 ms.  Throughput: 198.32 iter/sec.
Timings for 3360K FFT length (8 cores hyperthreaded, 1 worker):  2.16 ms.  Throughput: 463.70 iter/sec.
Timings for 3360K FFT length (8 cores hyperthreaded, 2 workers):  8.54,  8.53 ms.  Throughput: 234.36 iter/sec.
Timings for 3360K FFT length (8 cores hyperthreaded, 8 workers): 43.35, 41.38, 41.88, 41.34, 42.48, 41.29, 41.42, 41.06 ms.  Throughput: 191.55 iter/sec.
Timings for 3584K FFT length (8 cores, 1 worker):  2.53 ms.  Throughput: 395.49 iter/sec.
Timings for 3584K FFT length (8 cores, 2 workers):  8.79,  8.82 ms.  Throughput: 227.21 iter/sec.
Timings for 3584K FFT length (8 cores, 8 workers): 42.08, 42.27, 41.85, 42.15, 41.72, 41.83, 41.72, 41.74 ms.  Throughput: 190.85 iter/sec.
Timings for 3584K FFT length (8 cores hyperthreaded, 1 worker):  2.69 ms.  Throughput: 371.11 iter/sec.
Timings for 3584K FFT length (8 cores hyperthreaded, 2 workers):  9.40,  9.30 ms.  Throughput: 213.96 iter/sec.
Timings for 3584K FFT length (8 cores hyperthreaded, 8 workers): 45.64, 43.87, 43.73, 43.22, 44.67, 43.18, 43.73, 44.02 ms.  Throughput: 181.84 iter/sec.
Timings for 3840K FFT length (8 cores, 1 worker):  3.11 ms.  Throughput: 321.25 iter/sec.
Timings for 3840K FFT length (8 cores, 2 workers):  9.83,  9.75 ms.  Throughput: 204.28 iter/sec.
Timings for 3840K FFT length (8 cores, 8 workers): 45.17, 45.31, 45.17, 47.16, 45.44, 45.44, 45.58, 45.66 ms.  Throughput: 175.40 iter/sec.
Timings for 3840K FFT length (8 cores hyperthreaded, 1 worker):  3.35 ms.  Throughput: 298.94 iter/sec.
Timings for 3840K FFT length (8 cores hyperthreaded, 2 workers): 10.30, 10.13 ms.  Throughput: 195.75 iter/sec.
Timings for 3840K FFT length (8 cores hyperthreaded, 8 workers): 49.44, 46.58, 46.65, 46.50, 47.63, 46.42, 47.16, 46.76 ms.  Throughput: 169.77 iter/sec.
Timings for 4096K FFT length (8 cores, 1 worker):  3.38 ms.  Throughput: 295.98 iter/sec.
[Mon May 04 14:47:12 2020]
Timings for 4096K FFT length (8 cores, 2 workers): 11.39, 11.22 ms.  Throughput: 176.90 iter/sec.
Timings for 4096K FFT length (8 cores, 8 workers): 48.47, 49.21, 48.11, 49.45, 48.10, 48.10, 48.17, 48.77 ms.  Throughput: 164.80 iter/sec.
Timings for 4096K FFT length (8 cores hyperthreaded, 1 worker):  3.87 ms.  Throughput: 258.43 iter/sec.
Timings for 4096K FFT length (8 cores hyperthreaded, 2 workers): 11.26, 11.13 ms.  Throughput: 178.63 iter/sec.
Timings for 4096K FFT length (8 cores hyperthreaded, 8 workers): 52.77, 49.91, 49.53, 49.38, 51.00, 49.39, 50.20, 49.87 ms.  Throughput: 159.25 iter/sec.
Timings for 4480K FFT length (8 cores, 1 worker):  4.41 ms.  Throughput: 226.79 iter/sec.
Timings for 4480K FFT length (8 cores, 2 workers): 12.00, 12.00 ms.  Throughput: 166.68 iter/sec.
Timings for 4480K FFT length (8 cores, 8 workers): 53.99, 53.25, 53.06, 53.26, 52.98, 52.61, 53.03, 53.21 ms.  Throughput: 150.46 iter/sec.
Timings for 4480K FFT length (8 cores hyperthreaded, 1 worker):  4.95 ms.  Throughput: 202.01 iter/sec.
Timings for 4480K FFT length (8 cores hyperthreaded, 2 workers): 12.71, 12.59 ms.  Throughput: 158.10 iter/sec.
Timings for 4480K FFT length (8 cores hyperthreaded, 8 workers): 58.02, 55.16, 54.88, 54.61, 55.70, 54.42, 55.16, 55.22 ms.  Throughput: 144.46 iter/sec.
Timings for 4608K FFT length (8 cores, 1 worker):  4.69 ms.  Throughput: 213.16 iter/sec.
Timings for 4608K FFT length (8 cores, 2 workers): 12.51, 12.48 ms.  Throughput: 160.07 iter/sec.
Timings for 4608K FFT length (8 cores, 8 workers): 54.96, 55.43, 54.69, 55.00, 54.65, 54.97, 55.04, 54.92 ms.  Throughput: 145.57 iter/sec.
Timings for 4608K FFT length (8 cores hyperthreaded, 1 worker):  4.96 ms.  Throughput: 201.71 iter/sec.
Timings for 4608K FFT length (8 cores hyperthreaded, 2 workers): 13.04, 12.99 ms.  Throughput: 153.65 iter/sec.
Timings for 4608K FFT length (8 cores hyperthreaded, 8 workers): 58.60, 55.78, 55.93, 56.60, 58.08, 56.26, 56.26, 56.27 ms.  Throughput: 141.08 iter/sec.
Timings for 4800K FFT length (8 cores, 1 worker):  4.95 ms.  Throughput: 201.83 iter/sec.
Timings for 4800K FFT length (8 cores, 2 workers): 13.07, 12.96 ms.  Throughput: 153.68 iter/sec.
[Mon May 04 14:52:15 2020]
Timings for 4800K FFT length (8 cores, 8 workers): 57.29, 57.79, 57.05, 57.47, 57.42, 57.42, 57.54, 57.48 ms.  Throughput: 139.30 iter/sec.
Timings for 4800K FFT length (8 cores hyperthreaded, 1 worker):  5.21 ms.  Throughput: 191.86 iter/sec.
Timings for 4800K FFT length (8 cores hyperthreaded, 2 workers): 13.63, 13.59 ms.  Throughput: 146.97 iter/sec.
Timings for 4800K FFT length (8 cores hyperthreaded, 8 workers): 61.84, 58.62, 59.13, 58.19, 60.21, 58.34, 59.15, 58.51 ms.  Throughput: 135.07 iter/sec.
Timings for 5120K FFT length (8 cores, 1 worker):  5.59 ms.  Throughput: 178.98 iter/sec.
Timings for 5120K FFT length (8 cores, 2 workers): 14.20, 14.01 ms.  Throughput: 141.81 iter/sec.
Timings for 5120K FFT length (8 cores, 8 workers): 61.21, 61.62, 60.80, 61.24, 61.06, 61.05, 61.14, 61.12 ms.  Throughput: 130.82 iter/sec.
Timings for 5120K FFT length (8 cores hyperthreaded, 1 worker):  5.97 ms.  Throughput: 167.43 iter/sec.
Timings for 5120K FFT length (8 cores hyperthreaded, 2 workers): 14.79, 14.77 ms.  Throughput: 135.31 iter/sec.
Timings for 5120K FFT length (8 cores hyperthreaded, 8 workers): 65.52, 62.46, 62.47, 63.13, 64.12, 62.28, 62.98, 62.69 ms.  Throughput: 126.61 iter/sec.
Timings for 5376K FFT length (8 cores, 1 worker):  6.28 ms.  Throughput: 159.30 iter/sec.
Timings for 5376K FFT length (8 cores, 2 workers): 15.07, 15.07 ms.  Throughput: 132.69 iter/sec.
Timings for 5376K FFT length (8 cores, 8 workers): 64.34, 64.72, 63.97, 64.37, 64.38, 64.34, 64.33, 64.57 ms.  Throughput: 124.27 iter/sec.
Timings for 5376K FFT length (8 cores hyperthreaded, 1 worker):  6.67 ms.  Throughput: 150.03 iter/sec.
Timings for 5376K FFT length (8 cores hyperthreaded, 2 workers): 15.71, 15.59 ms.  Throughput: 127.78 iter/sec.
Timings for 5376K FFT length (8 cores hyperthreaded, 8 workers): 69.18, 65.63, 65.73, 65.61, 67.71, 65.22, 65.81, 65.54 ms.  Throughput: 120.70 iter/sec.
Timings for 5600K FFT length (8 cores, 1 worker):  6.69 ms.  Throughput: 149.58 iter/sec.
Timings for 5600K FFT length (8 cores, 2 workers): 16.04, 16.09 ms.  Throughput: 124.50 iter/sec.
Timings for 5600K FFT length (8 cores, 8 workers): 66.65, 67.52, 66.44, 66.96, 66.52, 66.42, 66.59, 66.44 ms.  Throughput: 119.96 iter/sec.
[Mon May 04 14:57:19 2020]
Timings for 5600K FFT length (8 cores hyperthreaded, 1 worker):  7.16 ms.  Throughput: 139.73 iter/sec.
Timings for 5600K FFT length (8 cores hyperthreaded, 2 workers): 16.48, 16.46 ms.  Throughput: 121.44 iter/sec.
Timings for 5600K FFT length (8 cores hyperthreaded, 8 workers): 74.48, 70.00, 68.22, 68.02, 70.46, 67.97, 69.64, 67.99 ms.  Throughput: 115.04 iter/sec.
Timings for 5760K FFT length (8 cores, 1 worker):  7.23 ms.  Throughput: 138.22 iter/sec.
Timings for 5760K FFT length (8 cores, 2 workers): 16.52, 16.48 ms.  Throughput: 121.25 iter/sec.
Timings for 5760K FFT length (8 cores, 8 workers): 69.28, 69.50, 69.02, 69.32, 69.27, 69.23, 69.27, 68.88 ms.  Throughput: 115.57 iter/sec.
Timings for 5760K FFT length (8 cores hyperthreaded, 1 worker):  7.94 ms.  Throughput: 125.97 iter/sec.
Timings for 5760K FFT length (8 cores hyperthreaded, 2 workers): 17.32, 17.27 ms.  Throughput: 115.64 iter/sec.
Timings for 5760K FFT length (8 cores hyperthreaded, 8 workers): 75.94, 71.74, 70.80, 70.58, 73.03, 70.65, 72.14, 71.28 ms.  Throughput: 111.14 iter/sec.
Timings for 6144K FFT length (8 cores, 1 worker):  7.69 ms.  Throughput: 130.03 iter/sec.
Timings for 6144K FFT length (8 cores, 2 workers): 17.43, 17.60 ms.  Throughput: 114.21 iter/sec.
Timings for 6144K FFT length (8 cores, 8 workers): 73.66, 74.15, 73.25, 73.75, 73.36, 73.36, 73.69, 73.73 ms.  Throughput: 108.67 iter/sec.
Timings for 6144K FFT length (8 cores hyperthreaded, 1 worker):  8.08 ms.  Throughput: 123.76 iter/sec.
Timings for 6144K FFT length (8 cores hyperthreaded, 2 workers): 18.52, 18.10 ms.  Throughput: 109.24 iter/sec.
Timings for 6144K FFT length (8 cores hyperthreaded, 8 workers): 78.84, 74.88, 75.70, 74.60, 77.34, 74.55, 76.13, 74.77 ms.  Throughput: 105.51 iter/sec.
Timings for 6400K FFT length (8 cores, 1 worker):  8.29 ms.  Throughput: 120.64 iter/sec.
Timings for 6400K FFT length (8 cores, 2 workers): 18.78, 18.78 ms.  Throughput: 106.49 iter/sec.
Timings for 6400K FFT length (8 cores, 8 workers): 77.37, 77.62, 76.74, 77.33, 77.14, 76.96, 77.31, 77.41 ms.  Throughput: 103.58 iter/sec.
Timings for 6400K FFT length (8 cores hyperthreaded, 1 worker):  8.55 ms.  Throughput: 116.98 iter/sec.
[Mon May 04 15:02:27 2020]
Timings for 6400K FFT length (8 cores hyperthreaded, 2 workers): 19.00, 18.86 ms.  Throughput: 105.64 iter/sec.
Timings for 6400K FFT length (8 cores hyperthreaded, 8 workers): 82.58, 78.49, 78.00, 77.47, 80.53, 77.58, 78.88, 77.99 ms.  Throughput: 101.39 iter/sec.
Timings for 6720K FFT length (8 cores, 1 worker):  8.81 ms.  Throughput: 113.46 iter/sec.
Timings for 6720K FFT length (8 cores, 2 workers): 19.35, 19.24 ms.  Throughput: 103.66 iter/sec.
Timings for 6720K FFT length (8 cores, 8 workers): 80.26, 80.53, 79.70, 80.31, 80.20, 79.48, 80.27, 80.27 ms.  Throughput: 99.84 iter/sec.
Timings for 6720K FFT length (8 cores hyperthreaded, 1 worker):  9.39 ms.  Throughput: 106.44 iter/sec.
Timings for 6720K FFT length (8 cores hyperthreaded, 2 workers): 20.20, 20.17 ms.  Throughput: 99.09 iter/sec.
Timings for 6720K FFT length (8 cores hyperthreaded, 8 workers): 87.80, 82.61, 81.73, 81.32, 84.14, 82.22, 82.74, 81.84 ms.  Throughput: 96.38 iter/sec.
Timings for 7168K FFT length (8 cores, 1 worker):  9.59 ms.  Throughput: 104.23 iter/sec.
Timings for 7168K FFT length (8 cores, 2 workers): 20.66, 20.56 ms.  Throughput: 97.03 iter/sec.
Timings for 7168K FFT length (8 cores, 8 workers): 88.03, 86.01, 85.41, 86.06, 85.50, 85.50, 85.87, 85.83 ms.  Throughput: 93.00 iter/sec.
Timings for 7168K FFT length (8 cores hyperthreaded, 1 worker): 10.03 ms.  Throughput: 99.74 iter/sec.
Timings for 7168K FFT length (8 cores hyperthreaded, 2 workers): 21.62, 21.61 ms.  Throughput: 92.53 iter/sec.
Timings for 7168K FFT length (8 cores hyperthreaded, 8 workers): 92.59, 88.47, 87.25, 87.26, 91.17, 86.84, 88.06, 87.37 ms.  Throughput: 90.31 iter/sec.
Timings for 7680K FFT length (8 cores, 1 worker): 10.64 ms.  Throughput: 93.99 iter/sec.
Timings for 7680K FFT length (8 cores, 2 workers): 22.48, 22.54 ms.  Throughput: 88.84 iter/sec.
Timings for 7680K FFT length (8 cores, 8 workers): 92.22, 93.02, 91.98, 92.49, 92.15, 91.94, 92.00, 91.72 ms.  Throughput: 86.78 iter/sec.
Timings for 7680K FFT length (8 cores hyperthreaded, 1 worker): 11.45 ms.  Throughput: 87.33 iter/sec.
Timings for 7680K FFT length (8 cores hyperthreaded, 2 workers): 23.41, 23.40 ms.  Throughput: 85.45 iter/sec.
[Mon May 04 15:07:39 2020]
Timings for 7680K FFT length (8 cores hyperthreaded, 8 workers): 101.05, 94.84, 94.87, 94.46, 96.86, 95.18, 95.95, 94.48 ms.  Throughput: 83.40 iter/sec.
Timings for 8000K FFT length (8 cores, 1 worker): 11.16 ms.  Throughput: 89.60 iter/sec.
Timings for 8000K FFT length (8 cores, 2 workers): 23.69, 23.78 ms.  Throughput: 84.26 iter/sec.
Timings for 8000K FFT length (8 cores, 8 workers): 98.10, 99.44, 97.46, 98.94, 98.47, 98.23, 98.60, 98.89 ms.  Throughput: 81.21 iter/sec.
Timings for 8000K FFT length (8 cores hyperthreaded, 1 worker): 11.56 ms.  Throughput: 86.50 iter/sec.
Timings for 8000K FFT length (8 cores hyperthreaded, 2 workers): 24.68, 24.44 ms.  Throughput: 81.44 iter/sec.
Timings for 8000K FFT length (8 cores hyperthreaded, 8 workers): 107.81, 99.28, 99.67, 97.19, 100.92, 97.75, 99.79, 99.41 ms.  Throughput: 79.89 iter/sec.
Timings for 8064K FFT length (8 cores, 1 worker): 11.32 ms.  Throughput: 88.34 iter/sec.
Timings for 8064K FFT length (8 cores, 2 workers): 23.69, 23.78 ms.  Throughput: 84.26 iter/sec.
Timings for 8064K FFT length (8 cores, 8 workers): 97.03, 97.16, 96.80, 97.29, 96.86, 96.87, 97.07, 97.18 ms.  Throughput: 82.45 iter/sec.
Timings for 8064K FFT length (8 cores hyperthreaded, 1 worker): 11.81 ms.  Throughput: 84.65 iter/sec.
Timings for 8064K FFT length (8 cores hyperthreaded, 2 workers): 24.56, 24.52 ms.  Throughput: 81.49 iter/sec.
Timings for 8064K FFT length (8 cores hyperthreaded, 8 workers): 103.89, 98.97, 98.90, 98.77, 101.96, 98.84, 99.58, 98.78 ms.  Throughput: 80.06 iter/sec.
Timings for 8192K FFT length (8 cores, 1 worker): 11.47 ms.  Throughput: 87.17 iter/sec.
Timings for 8192K FFT length (8 cores, 2 workers): 24.11, 24.10 ms.  Throughput: 82.99 iter/sec.
Timings for 8192K FFT length (8 cores, 8 workers): 98.64, 99.15, 98.13, 98.72, 98.21, 98.36, 98.44, 98.69 ms.  Throughput: 81.18 iter/sec.
Timings for 8192K FFT length (8 cores hyperthreaded, 1 worker): 12.00 ms.  Throughput: 83.33 iter/sec.
Timings for 8192K FFT length (8 cores hyperthreaded, 2 workers): 25.09, 24.91 ms.  Throughput: 80.00 iter/sec.
Timings for 8192K FFT length (8 cores hyperthreaded, 8 workers): 106.03, 101.67, 100.12, 99.97, 103.77, 99.69, 101.85, 100.41 ms.  Throughput: 78.70 iter/sec.

FFT-Timings:

Code:

AMD Ryzen 7 3700X 8-Core Processor             
CPU speed: 4293.85 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: 2x16 MB
L1 cache line size: 64 bytes, L2 cache line size: 64 bytes
Machine topology as determined by hwloc library:
 Machine#0 (total=14412868KB, Backend=Windows, hwlocVersion=2.0.4, ProcessName=prime95.exe)
  Package (total=14412868KB, CPUVendor=AuthenticAMD, CPUFamilyNumber=23, CPUModelNumber=113, CPUModel="AMD Ryzen 7 3700X 8-Core Processor             ", CPUStepping=0)
    L3 (size=16384KB, 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)
    L3 (size=16384KB, 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: 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 29.8, RdtscTiming=1
Timing FFTs using 8 threads on 8 cores.
Best time for 2048K FFT length: 1.205 ms., avg: 1.224 ms.
Best time for 2240K FFT length: 1.279 ms., avg: 1.314 ms.
Best time for 2304K FFT length: 1.320 ms., avg: 1.342 ms.
Best time for 2400K FFT length: 1.382 ms., avg: 1.411 ms.
Best time for 2560K FFT length: 1.498 ms., avg: 1.532 ms.
Best time for 2688K FFT length: 1.597 ms., avg: 1.626 ms.
Best time for 2800K FFT length: 1.880 ms., avg: 1.915 ms.
Best time for 2880K FFT length: 1.699 ms., avg: 1.734 ms.
Best time for 3072K FFT length: 1.779 ms., avg: 1.847 ms.
Best time for 3200K FFT length: 2.058 ms., avg: 2.133 ms.
Best time for 3360K FFT length: 2.095 ms., avg: 2.167 ms.
Best time for 3584K FFT length: 2.387 ms., avg: 2.501 ms.
Best time for 3840K FFT length: 3.044 ms., avg: 3.214 ms.
Best time for 4096K FFT length: 3.236 ms., avg: 3.406 ms.
Best time for 4480K FFT length: 4.265 ms., avg: 4.382 ms.
Best time for 4608K FFT length: 4.498 ms., avg: 4.593 ms.
Best time for 4800K FFT length: 4.820 ms., avg: 4.943 ms.
Best time for 5120K FFT length: 5.556 ms., avg: 5.815 ms.
Best time for 5376K FFT length: 6.160 ms., avg: 6.252 ms.
Best time for 5600K FFT length: 6.604 ms., avg: 6.743 ms.
Best time for 5760K FFT length: 7.095 ms., avg: 7.262 ms.
Best time for 6144K FFT length: 7.617 ms., avg: 7.727 ms.
Best time for 6400K FFT length: 8.140 ms., avg: 8.287 ms.
Best time for 6720K FFT length: 8.686 ms., avg: 8.826 ms.
Best time for 7168K FFT length: 9.463 ms., avg: 9.606 ms.
Best time for 7680K FFT length: 10.560 ms., avg: 10.681 ms.
Best time for 8000K FFT length: 11.068 ms., avg: 11.191 ms.
Best time for 8064K FFT length: 11.207 ms., avg: 11.327 ms.
Best time for 8192K FFT length: 11.377 ms., avg: 11.488 ms.

TF benchmark:

Code:

AMD Ryzen 7 3700X 8-Core Processor             
CPU speed: 4298.80 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: 2x16 MB
L1 cache line size: 64 bytes, L2 cache line size: 64 bytes
Machine topology as determined by hwloc library:
 Machine#0 (total=14412868KB, Backend=Windows, hwlocVersion=2.0.4, ProcessName=prime95.exe)
  Package (total=14412868KB, CPUVendor=AuthenticAMD, CPUFamilyNumber=23, CPUModelNumber=113, CPUModel="AMD Ryzen 7 3700X 8-Core Processor             ", CPUStepping=0)
    L3 (size=16384KB, 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)
    L3 (size=16384KB, 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: 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 29.8, RdtscTiming=1
Best time for 61 bit trial factors: 0.691 ms.
Best time for 62 bit trial factors: 0.696 ms.
Best time for 63 bit trial factors: 0.703 ms.
Best time for 64 bit trial factors: 0.790 ms.
Best time for 65 bit trial factors: 0.691 ms.
Best time for 66 bit trial factors: 0.678 ms.
Best time for 67 bit trial factors: 0.685 ms.
Best time for 75 bit trial factors: 0.676 ms.
Best time for 76 bit trial factors: 0.707 ms.
Best time for 77 bit trial factors: 0.677 ms.

axn · 2020-05-04, 15:08

Quote:

Originally Posted by moebius

Throughput-Test:

AMD Ryzen 7 3700X 8-Core Processor

Timings for 5760K FFT length (8 cores, 1 worker): 7.23 ms. Throughput: 138.22 iter/sec.

Best time for 5760K FFT length: 7.095 ms., avg: 7.262 ms.

Something's wrong. On my 6 core 3600, I'm getting

Code:

FFTlen=5760K, Type=3, Arch=4, Pass1=1280, Pass2=4608, clm=2 (6 cores, 1 worker):  4.85 ms.  Throughput: 206.36 iter/sec.

I do have 2x16 GB DDR4 3200 RAM, but that's not enough to explain the difference.

EDIT:- Have you enabled XMP in BIOS? If not, the RAM might be running at slower than rated speed.

moebius · 2020-05-04, 17:43

Quote:

Originally Posted by axn

Something's wrong.

You're right, I had not turned on XAMP in the X470 mainboard bios, the G SKILL Aegis (also cheap RAM) only ran with a CL 15 profile.

FFT-Timings:

Code:

Prime95 64-bit version 29.8, RdtscTiming=1
Timing FFTs using 8 threads on 8 cores.
Best time for 2048K FFT length: 1.069 ms., avg: 1.093 ms.
Best time for 2240K FFT length: 1.190 ms., avg: 1.313 ms.
Best time for 2304K FFT length: 1.161 ms., avg: 1.179 ms.
Best time for 2400K FFT length: 1.247 ms., avg: 1.266 ms.
Best time for 2560K FFT length: 1.328 ms., avg: 1.359 ms.
Best time for 2688K FFT length: 1.413 ms., avg: 1.436 ms.
Best time for 2800K FFT length: 1.424 ms., avg: 1.447 ms.
Best time for 2880K FFT length: 1.472 ms., avg: 1.496 ms.
Best time for 3072K FFT length: 1.568 ms., avg: 1.608 ms.
Best time for 3200K FFT length: 1.725 ms., avg: 1.764 ms.
Best time for 3360K FFT length: 1.838 ms., avg: 1.879 ms.
Best time for 3584K FFT length: 2.057 ms., avg: 2.138 ms.
Best time for 3840K FFT length: 2.630 ms., avg: 2.806 ms.
Best time for 4096K FFT length: 2.897 ms., avg: 2.999 ms.
Best time for 4480K FFT length: 3.767 ms., avg: 3.869 ms.
Best time for 4608K FFT length: 4.117 ms., avg: 4.202 ms.
Best time for 4800K FFT length: 4.418 ms., avg: 4.490 ms.
Best time for 5120K FFT length: 5.082 ms., avg: 5.159 ms.
Best time for 5376K FFT length: 5.726 ms., avg: 5.792 ms.
Best time for 5600K FFT length: 6.116 ms., avg: 6.194 ms.
Best time for 5760K FFT length: 6.585 ms., avg: 6.698 ms.
Best time for 6144K FFT length: 7.127 ms., avg: 7.210 ms.
Best time for 6400K FFT length: 7.666 ms., avg: 7.755 ms.
Best time for 6720K FFT length: 8.044 ms., avg: 8.157 ms.
Best time for 7168K FFT length: 8.893 ms., avg: 8.992 ms.
Best time for 7680K FFT length: 9.871 ms., avg: 9.960 ms.
Best time for 8000K FFT length: 10.394 ms., avg: 10.492 ms.
Best time for 8064K FFT length: 10.519 ms., avg: 10.635 ms.
Best time for 8192K FFT length: 10.665 ms., avg: 10.798 ms.

Throughput-benchmarks for all 5760 FFT implemantations

Code:

Prime95 64-bit version 29.8, RdtscTiming=1
FFTlen=5760K, Type=3, Arch=4, Pass1=384, Pass2=15360, clm=4 (8 cores, 1 worker):  6.65 ms.  Throughput: 150.33 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=384, Pass2=15360, clm=4 (8 cores hyperthreaded, 1 worker):  7.23 ms.  Throughput: 138.31 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=384, Pass2=15360, clm=2 (8 cores, 1 worker):  6.63 ms.  Throughput: 150.82 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=384, Pass2=15360, clm=2 (8 cores hyperthreaded, 1 worker):  7.05 ms.  Throughput: 141.85 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=384, Pass2=15360, clm=1 (8 cores, 1 worker):  6.88 ms.  Throughput: 145.26 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=384, Pass2=15360, clm=1 (8 cores hyperthreaded, 1 worker):  7.24 ms.  Throughput: 138.09 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=640, Pass2=9216, clm=4 (8 cores, 1 worker):  6.60 ms.  Throughput: 151.56 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=640, Pass2=9216, clm=4 (8 cores hyperthreaded, 1 worker):  7.28 ms.  Throughput: 137.40 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=640, Pass2=9216, clm=2 (8 cores, 1 worker):  6.47 ms.  Throughput: 154.45 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=640, Pass2=9216, clm=2 (8 cores hyperthreaded, 1 worker):  7.07 ms.  Throughput: 141.51 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=640, Pass2=9216, clm=1 (8 cores, 1 worker):  6.61 ms.  Throughput: 151.18 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=640, Pass2=9216, clm=1 (8 cores hyperthreaded, 1 worker):  7.04 ms.  Throughput: 141.96 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=768, Pass2=7680, clm=4 (8 cores, 1 worker):  6.70 ms.  Throughput: 149.27 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=768, Pass2=7680, clm=4 (8 cores hyperthreaded, 1 worker):  7.47 ms.  Throughput: 133.95 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=768, Pass2=7680, clm=2 (8 cores, 1 worker):  6.55 ms.  Throughput: 152.75 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=768, Pass2=7680, clm=2 (8 cores hyperthreaded, 1 worker):  7.15 ms.  Throughput: 139.76 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=768, Pass2=7680, clm=1 (8 cores, 1 worker):  6.68 ms.  Throughput: 149.76 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=768, Pass2=7680, clm=1 (8 cores hyperthreaded, 1 worker):  7.11 ms.  Throughput: 140.59 iter/sec.
[Mon May 04 19:20:51 2020]
FFTlen=5760K, Type=3, Arch=4, Pass1=1280, Pass2=4608, clm=4 (8 cores, 1 worker):  6.82 ms.  Throughput: 146.67 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1280, Pass2=4608, clm=4 (8 cores hyperthreaded, 1 worker):  7.62 ms.  Throughput: 131.29 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1280, Pass2=4608, clm=2 (8 cores, 1 worker):  6.56 ms.  Throughput: 152.33 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1280, Pass2=4608, clm=2 (8 cores hyperthreaded, 1 worker):  7.20 ms.  Throughput: 138.96 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1280, Pass2=4608, clm=1 (8 cores, 1 worker):  6.68 ms.  Throughput: 149.70 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1280, Pass2=4608, clm=1 (8 cores hyperthreaded, 1 worker):  7.08 ms.  Throughput: 141.16 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1536, Pass2=3840, clm=4 (8 cores, 1 worker):  6.87 ms.  Throughput: 145.61 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1536, Pass2=3840, clm=4 (8 cores hyperthreaded, 1 worker):  7.72 ms.  Throughput: 129.57 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1536, Pass2=3840, clm=2 (8 cores, 1 worker):  6.49 ms.  Throughput: 154.06 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1536, Pass2=3840, clm=2 (8 cores hyperthreaded, 1 worker):  7.15 ms.  Throughput: 139.87 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1536, Pass2=3840, clm=1 (8 cores, 1 worker):  6.61 ms.  Throughput: 151.38 iter/sec.
FFTlen=5760K, Type=3, Arch=4, Pass1=1536, Pass2=3840, clm=1 (8 cores hyperthreaded, 1 worker):  6.97 ms.  Throughput: 143.46 iter/sec.

xx005fs · 2020-05-06, 21:40

CPU clockspeed locked at 4.2GHz, running DDR4 3733 with Dual channel and Dual Rank memory, FCLK at 1866MHz.

Code:

Prime95 64-bit version 29.8, RdtscTiming=1
Timings for 2048K FFT length (12 cores, 1 worker):  1.15 ms.  Throughput: 868.56 iter/sec.
Timings for 2048K FFT length (12 cores, 2 workers):  1.42,  1.43 ms.  Throughput: 1406.20 iter/sec.
Timings for 2048K FFT length (12 cores, 4 workers):  2.39,  2.39,  2.37,  2.40 ms.  Throughput: 1674.66 iter/sec.
Timings for 2048K FFT length (12 cores, 12 workers): 16.20, 16.20, 16.15, 16.22, 16.25, 16.25, 15.97, 15.87, 15.97, 15.70, 15.65, 15.87 ms.  Throughput: 748.96 iter/sec.
Timings for 2048K FFT length (12 cores hyperthreaded, 1 worker):  1.12 ms.  Throughput: 891.90 iter/sec.
Timings for 2048K FFT length (12 cores hyperthreaded, 2 workers):  1.34,  1.34 ms.  Throughput: 1494.43 iter/sec.
Timings for 2048K FFT length (12 cores hyperthreaded, 4 workers):  2.45,  2.44,  2.44,  2.45 ms.  Throughput: 1635.32 iter/sec.
Timings for 2048K FFT length (12 cores hyperthreaded, 12 workers): 16.49, 15.96, 15.96, 16.27, 16.35, 16.30, 16.21, 16.18, 16.18, 16.03, 16.33, 16.04 ms.  Throughput: 741.18 iter/sec.
Timings for 2240K FFT length (12 cores, 1 worker):  1.00 ms.  Throughput: 1001.96 iter/sec.
Timings for 2240K FFT length (12 cores, 2 workers):  1.54,  1.56 ms.  Throughput: 1289.98 iter/sec.
Timings for 2240K FFT length (12 cores, 4 workers):  3.02,  3.05,  3.05,  3.05 ms.  Throughput: 1314.66 iter/sec.
Timings for 2240K FFT length (12 cores, 12 workers): 17.95, 17.92, 17.88, 17.92, 17.92, 17.89, 17.81, 17.81, 17.81, 17.58, 17.80, 17.76 ms.  Throughput: 672.76 iter/sec.
Timings for 2240K FFT length (12 cores hyperthreaded, 1 worker):  0.99 ms.  Throughput: 1013.82 iter/sec.
Timings for 2240K FFT length (12 cores hyperthreaded, 2 workers):  1.37,  1.37 ms.  Throughput: 1462.48 iter/sec.
Timings for 2240K FFT length (12 cores hyperthreaded, 4 workers):  3.07,  3.03,  3.01,  3.04 ms.  Throughput: 1316.54 iter/sec.
Timings for 2240K FFT length (12 cores hyperthreaded, 12 workers): 18.43, 17.98, 17.80, 18.02, 18.22, 18.12, 18.07, 17.94, 17.97, 17.56, 18.58, 17.50 ms.  Throughput: 666.27 iter/sec.
Timings for 2304K FFT length (12 cores, 1 worker):  1.02 ms.  Throughput: 978.43 iter/sec.
Timings for 2304K FFT length (12 cores, 2 workers):  1.53,  1.53 ms.  Throughput: 1307.58 iter/sec.
Timings for 2304K FFT length (12 cores, 4 workers):  3.33,  3.35,  3.24,  3.33 ms.  Throughput: 1207.35 iter/sec.
Timings for 2304K FFT length (12 cores, 12 workers): 18.64, 18.64, 18.58, 18.56, 18.46, 18.56, 18.09, 18.45, 18.45, 18.21, 18.21, 18.36 ms.  Throughput: 651.00 iter/sec.
Timings for 2304K FFT length (12 cores hyperthreaded, 1 worker):  1.28 ms.  Throughput: 780.58 iter/sec.
Timings for 2304K FFT length (12 cores hyperthreaded, 2 workers):  1.36,  1.37 ms.  Throughput: 1465.32 iter/sec.
Timings for 2304K FFT length (12 cores hyperthreaded, 4 workers):  3.37,  3.36,  3.36,  3.36 ms.  Throughput: 1189.00 iter/sec.
Timings for 2304K FFT length (12 cores hyperthreaded, 12 workers): 18.96, 18.35, 18.48, 18.62, 18.69, 18.65, 18.33, 18.29, 18.82, 18.27, 18.92, 18.27 ms.  Throughput: 646.90 iter/sec.
Timings for 2400K FFT length (12 cores, 1 worker):  1.08 ms.  Throughput: 929.98 iter/sec.
Timings for 2400K FFT length (12 cores, 2 workers):  1.66,  1.64 ms.  Throughput: 1212.76 iter/sec.
Timings for 2400K FFT length (12 cores, 4 workers):  3.87,  3.80,  3.84,  3.82 ms.  Throughput: 1044.07 iter/sec.
Timings for 2400K FFT length (12 cores, 12 workers): 19.64, 19.64, 19.64, 19.38, 19.59, 19.59, 19.41, 19.45, 19.41, 19.41, 19.41, 19.41 ms.  Throughput: 615.45 iter/sec.
[Wed May 06 12:42:01 2020]
Timings for 2400K FFT length (12 cores hyperthreaded, 1 worker):  1.20 ms.  Throughput: 835.68 iter/sec.
Timings for 2400K FFT length (12 cores hyperthreaded, 2 workers):  1.42,  1.43 ms.  Throughput: 1407.33 iter/sec.
Timings for 2400K FFT length (12 cores hyperthreaded, 4 workers):  3.94,  3.91,  3.91,  3.94 ms.  Throughput: 1019.26 iter/sec.
Timings for 2400K FFT length (12 cores hyperthreaded, 12 workers): 19.75, 19.46, 19.44, 19.47, 19.56, 19.47, 19.57, 19.13, 19.57, 18.98, 20.08, 19.15 ms.  Throughput: 616.44 iter/sec.
Timings for 2560K FFT length (12 cores, 1 worker):  1.15 ms.  Throughput: 871.13 iter/sec.
Timings for 2560K FFT length (12 cores, 2 workers):  1.77,  1.78 ms.  Throughput: 1125.00 iter/sec.
Timings for 2560K FFT length (12 cores, 4 workers):  4.20,  4.19,  4.13,  4.15 ms.  Throughput: 960.33 iter/sec.
Timings for 2560K FFT length (12 cores, 12 workers): 20.48, 20.56, 20.56, 20.38, 20.33, 20.13, 20.28, 20.24, 20.28, 19.99, 20.19, 20.19 ms.  Throughput: 591.18 iter/sec.
Timings for 2560K FFT length (12 cores hyperthreaded, 1 worker):  1.11 ms.  Throughput: 901.22 iter/sec.
Timings for 2560K FFT length (12 cores hyperthreaded, 2 workers):  1.57,  1.57 ms.  Throughput: 1271.00 iter/sec.
Timings for 2560K FFT length (12 cores hyperthreaded, 4 workers):  4.40,  4.35,  4.27,  4.32 ms.  Throughput: 923.19 iter/sec.
Timings for 2560K FFT length (12 cores hyperthreaded, 12 workers): 20.96, 20.91, 20.30, 20.53, 20.57, 20.76, 20.24, 20.62, 20.57, 20.00, 20.88, 20.58 ms.  Throughput: 583.33 iter/sec.
Timings for 2688K FFT length (12 cores, 1 worker):  1.16 ms.  Throughput: 860.66 iter/sec.
Timings for 2688K FFT length (12 cores, 2 workers):  1.83,  1.83 ms.  Throughput: 1093.58 iter/sec.
Timings for 2688K FFT length (12 cores, 4 workers):  4.98,  5.00,  4.93,  4.82 ms.  Throughput: 811.38 iter/sec.
Timings for 2688K FFT length (12 cores, 12 workers): 21.81, 21.57, 21.81, 21.92, 21.92, 21.92, 21.71, 21.41, 21.71, 21.40, 21.20, 21.63 ms.  Throughput: 553.85 iter/sec.
Timings for 2688K FFT length (12 cores hyperthreaded, 1 worker):  1.16 ms.  Throughput: 859.74 iter/sec.
Timings for 2688K FFT length (12 cores hyperthreaded, 2 workers):  1.62,  1.62 ms.  Throughput: 1234.22 iter/sec.
Timings for 2688K FFT length (12 cores hyperthreaded, 4 workers):  5.14,  5.06,  5.06,  5.06 ms.  Throughput: 786.86 iter/sec.
Timings for 2688K FFT length (12 cores hyperthreaded, 12 workers): 22.31, 21.58, 21.72, 21.99, 22.04, 21.64, 21.66, 22.00, 21.65, 21.42, 22.38, 21.41 ms.  Throughput: 550.13 iter/sec.
Timings for 2800K FFT length (12 cores, 1 worker):  1.66 ms.  Throughput: 601.34 iter/sec.
Timings for 2800K FFT length (12 cores, 2 workers):  1.93,  1.93 ms.  Throughput: 1036.43 iter/sec.
Timings for 2800K FFT length (12 cores, 4 workers):  5.64,  5.58,  5.53,  5.56 ms.  Throughput: 717.19 iter/sec.
Timings for 2800K FFT length (12 cores, 12 workers): 23.01, 23.02, 22.86, 22.97, 22.86, 22.97, 22.81, 22.43, 22.70, 22.82, 22.82, 22.65 ms.  Throughput: 525.71 iter/sec.
Timings for 2800K FFT length (12 cores hyperthreaded, 1 worker):  1.63 ms.  Throughput: 613.38 iter/sec.
Timings for 2800K FFT length (12 cores hyperthreaded, 2 workers):  1.82,  1.99 ms.  Throughput: 1051.50 iter/sec.
Timings for 2800K FFT length (12 cores hyperthreaded, 4 workers):  5.78,  5.65,  5.73,  5.50 ms.  Throughput: 706.21 iter/sec.
Timings for 2800K FFT length (12 cores hyperthreaded, 12 workers): 24.29, 22.44, 23.28, 22.72, 23.20, 23.96, 22.65, 23.22, 23.71, 23.38, 24.23, 21.98 ms.  Throughput: 516.47 iter/sec.
Timings for 2880K FFT length (12 cores, 1 worker):  1.25 ms.  Throughput: 801.81 iter/sec.
[Wed May 06 12:47:06 2020]
Timings for 2880K FFT length (12 cores, 2 workers):  2.04,  2.04 ms.  Throughput: 980.36 iter/sec.
Timings for 2880K FFT length (12 cores, 4 workers):  5.88,  5.80,  5.86,  5.84 ms.  Throughput: 684.55 iter/sec.
Timings for 2880K FFT length (12 cores, 12 workers): 23.70, 23.70, 23.70, 23.55, 23.56, 23.36, 23.53, 23.53, 23.53, 22.88, 22.99, 23.54 ms.  Throughput: 511.44 iter/sec.
Timings for 2880K FFT length (12 cores hyperthreaded, 1 worker):  1.27 ms.  Throughput: 785.10 iter/sec.
Timings for 2880K FFT length (12 cores hyperthreaded, 2 workers):  1.72,  1.69 ms.  Throughput: 1174.38 iter/sec.
Timings for 2880K FFT length (12 cores hyperthreaded, 4 workers):  5.95,  5.76,  5.93,  5.85 ms.  Throughput: 681.33 iter/sec.
Timings for 2880K FFT length (12 cores hyperthreaded, 12 workers): 24.20, 23.36, 23.49, 23.40, 23.91, 23.60, 23.50, 23.26, 23.68, 22.91, 24.43, 23.23 ms.  Throughput: 509.02 iter/sec.
Timings for 3072K FFT length (12 cores, 1 worker):  1.34 ms.  Throughput: 744.55 iter/sec.
Timings for 3072K FFT length (12 cores, 2 workers):  2.04,  2.04 ms.  Throughput: 979.83 iter/sec.
Timings for 3072K FFT length (12 cores, 4 workers):  6.65,  6.54,  6.53,  6.51 ms.  Throughput: 610.20 iter/sec.
Timings for 3072K FFT length (12 cores, 12 workers): 25.16, 25.07, 24.87, 25.22, 25.21, 25.08, 24.98, 25.02, 24.91, 24.83, 24.83, 25.06 ms.  Throughput: 479.62 iter/sec.
Timings for 3072K FFT length (12 cores hyperthreaded, 1 worker):  1.30 ms.  Throughput: 767.92 iter/sec.
Timings for 3072K FFT length (12 cores hyperthreaded, 2 workers):  1.87,  1.88 ms.  Throughput: 1067.22 iter/sec.
Timings for 3072K FFT length (12 cores hyperthreaded, 4 workers):  6.75,  6.63,  6.57,  6.64 ms.  Throughput: 601.77 iter/sec.
Timings for 3072K FFT length (12 cores hyperthreaded, 12 workers): 26.21, 24.93, 25.18, 25.27, 25.55, 25.35, 25.15, 25.14, 25.21, 25.03, 25.69, 25.03 ms.  Throughput: 474.17 iter/sec.
Timings for 3200K FFT length (12 cores, 1 worker):  1.43 ms.  Throughput: 701.25 iter/sec.
Timings for 3200K FFT length (12 cores, 2 workers):  2.25,  2.20 ms.  Throughput: 898.79 iter/sec.
Timings for 3200K FFT length (12 cores, 4 workers):  6.81,  6.88,  6.77,  6.82 ms.  Throughput: 586.50 iter/sec.
Timings for 3200K FFT length (12 cores, 12 workers): 25.80, 25.74, 25.32, 26.23, 26.23, 26.23, 26.02, 26.02, 26.02, 25.49, 25.62, 25.69 ms.  Throughput: 463.94 iter/sec.
Timings for 3200K FFT length (12 cores hyperthreaded, 1 worker):  1.36 ms.  Throughput: 736.47 iter/sec.
Timings for 3200K FFT length (12 cores hyperthreaded, 2 workers):  1.95,  1.95 ms.  Throughput: 1023.69 iter/sec.
Timings for 3200K FFT length (12 cores hyperthreaded, 4 workers):  6.98,  7.25,  6.92,  6.99 ms.  Throughput: 568.90 iter/sec.
Timings for 3200K FFT length (12 cores hyperthreaded, 12 workers): 27.34, 25.74, 25.74, 26.33, 26.44, 25.83, 26.25, 25.56, 26.26, 25.67, 26.57, 25.93 ms.  Throughput: 459.24 iter/sec.
Timings for 3360K FFT length (12 cores, 1 worker):  1.46 ms.  Throughput: 686.03 iter/sec.
Timings for 3360K FFT length (12 cores, 2 workers):  2.39,  2.37 ms.  Throughput: 840.18 iter/sec.
Timings for 3360K FFT length (12 cores, 4 workers):  7.51,  7.69,  7.45,  7.45 ms.  Throughput: 531.51 iter/sec.
Timings for 3360K FFT length (12 cores, 12 workers): 27.63, 27.50, 27.26, 27.74, 27.74, 27.74, 27.27, 27.27, 27.42, 27.04, 27.47, 27.48 ms.  Throughput: 436.98 iter/sec.
Timings for 3360K FFT length (12 cores hyperthreaded, 1 worker):  1.39 ms.  Throughput: 719.35 iter/sec.
Timings for 3360K FFT length (12 cores hyperthreaded, 2 workers):  2.07,  2.08 ms.  Throughput: 963.41 iter/sec.
[Wed May 06 12:52:14 2020]
Timings for 3360K FFT length (12 cores hyperthreaded, 4 workers):  7.70,  7.55,  7.53,  7.53 ms.  Throughput: 527.72 iter/sec.
Timings for 3360K FFT length (12 cores hyperthreaded, 12 workers): 28.39, 27.60, 27.57, 27.91, 28.02, 27.19, 27.00, 27.12, 28.80, 27.21, 28.29, 27.29 ms.  Throughput: 433.40 iter/sec.
Timings for 3584K FFT length (12 cores, 1 worker):  1.55 ms.  Throughput: 643.82 iter/sec.
Timings for 3584K FFT length (12 cores, 2 workers):  2.48,  2.57 ms.  Throughput: 791.73 iter/sec.
Timings for 3584K FFT length (12 cores, 4 workers):  8.29,  8.29,  8.12,  8.29 ms.  Throughput: 485.06 iter/sec.
Timings for 3584K FFT length (12 cores, 12 workers): 29.52, 29.42, 29.15, 29.48, 29.48, 29.26, 29.10, 29.10, 29.39, 29.08, 28.73, 29.38 ms.  Throughput: 410.17 iter/sec.
Timings for 3584K FFT length (12 cores hyperthreaded, 1 worker):  1.48 ms.  Throughput: 675.07 iter/sec.
Timings for 3584K FFT length (12 cores hyperthreaded, 2 workers):  2.52,  2.52 ms.  Throughput: 793.59 iter/sec.
Timings for 3584K FFT length (12 cores hyperthreaded, 4 workers):  8.42,  8.35,  8.29,  8.35 ms.  Throughput: 478.83 iter/sec.
Timings for 3584K FFT length (12 cores hyperthreaded, 12 workers): 30.38, 29.40, 29.56, 29.76, 29.86, 29.50, 29.51, 29.60, 29.52, 29.30, 30.45, 28.94 ms.  Throughput: 404.83 iter/sec.
Timings for 3840K FFT length (12 cores, 1 worker):  1.62 ms.  Throughput: 615.43 iter/sec.
Timings for 3840K FFT length (12 cores, 2 workers):  2.75,  2.75 ms.  Throughput: 725.97 iter/sec.
Timings for 3840K FFT length (12 cores, 4 workers):  8.95,  8.90,  8.88,  8.89 ms.  Throughput: 449.21 iter/sec.
Timings for 3840K FFT length (12 cores, 12 workers): 31.25, 31.23, 31.23, 31.36, 31.36, 31.26, 31.26, 31.26, 31.26, 30.92, 30.93, 30.94 ms.  Throughput: 384.77 iter/sec.
Timings for 3840K FFT length (12 cores hyperthreaded, 1 worker):  1.57 ms.  Throughput: 635.79 iter/sec.
Timings for 3840K FFT length (12 cores hyperthreaded, 2 workers):  2.97,  2.97 ms.  Throughput: 674.03 iter/sec.
Timings for 3840K FFT length (12 cores hyperthreaded, 4 workers):  9.14,  9.01,  8.99,  9.06 ms.  Throughput: 442.08 iter/sec.
Timings for 3840K FFT length (12 cores hyperthreaded, 12 workers): 32.35, 31.22, 31.07, 31.36, 31.62, 31.47, 31.40, 31.39, 31.39, 30.96, 32.10, 30.80 ms.  Throughput: 381.91 iter/sec.
Timings for 4096K FFT length (12 cores, 1 worker):  1.74 ms.  Throughput: 576.07 iter/sec.
Timings for 4096K FFT length (12 cores, 2 workers):  3.33,  3.30 ms.  Throughput: 603.44 iter/sec.
Timings for 4096K FFT length (12 cores, 4 workers):  9.93,  9.84,  9.87,  9.76 ms.  Throughput: 406.00 iter/sec.
Timings for 4096K FFT length (12 cores, 12 workers): 33.66, 33.82, 33.41, 33.78, 33.53, 33.40, 33.41, 33.08, 33.55, 33.32, 33.06, 33.76 ms.  Throughput: 358.42 iter/sec.
Timings for 4096K FFT length (12 cores hyperthreaded, 1 worker):  1.68 ms.  Throughput: 596.51 iter/sec.
Timings for 4096K FFT length (12 cores hyperthreaded, 2 workers):  3.83,  3.83 ms.  Throughput: 522.63 iter/sec.
Timings for 4096K FFT length (12 cores hyperthreaded, 4 workers): 10.13, 10.04,  9.98, 10.01 ms.  Throughput: 398.52 iter/sec.
Timings for 4096K FFT length (12 cores hyperthreaded, 12 workers): 34.88, 33.72, 33.71, 33.74, 34.72, 33.63, 33.78, 33.53, 34.02, 33.75, 34.33, 33.39 ms.  Throughput: 353.68 iter/sec.
Timings for 4480K FFT length (12 cores, 1 worker):  1.87 ms.  Throughput: 535.64 iter/sec.
Timings for 4480K FFT length (12 cores, 2 workers):  4.32,  4.34 ms.  Throughput: 462.06 iter/sec.
Timings for 4480K FFT length (12 cores, 4 workers): 11.33, 11.25, 11.22, 11.17 ms.  Throughput: 355.83 iter/sec.
[Wed May 06 12:57:25 2020]
Timings for 4480K FFT length (12 cores, 12 workers): 37.51, 37.36, 36.97, 37.35, 37.26, 37.24, 37.04, 36.83, 37.11, 37.02, 36.85, 37.17 ms.  Throughput: 323.10 iter/sec.
Timings for 4480K FFT length (12 cores hyperthreaded, 1 worker):  1.83 ms.  Throughput: 546.92 iter/sec.
Timings for 4480K FFT length (12 cores hyperthreaded, 2 workers):  4.69,  4.72 ms.  Throughput: 425.18 iter/sec.
Timings for 4480K FFT length (12 cores hyperthreaded, 4 workers): 11.56, 11.43, 11.41, 11.45 ms.  Throughput: 348.92 iter/sec.
Timings for 4480K FFT length (12 cores hyperthreaded, 12 workers): 38.53, 37.54, 37.30, 37.34, 38.28, 37.54, 37.06, 37.15, 38.10, 37.04, 38.49, 36.90 ms.  Throughput: 319.17 iter/sec.
Timings for 4608K FFT length (12 cores, 1 worker):  1.91 ms.  Throughput: 522.93 iter/sec.
Timings for 4608K FFT length (12 cores, 2 workers):  4.14,  4.25 ms.  Throughput: 477.04 iter/sec.
Timings for 4608K FFT length (12 cores, 4 workers): 11.34, 11.30, 11.32, 11.32 ms.  Throughput: 353.37 iter/sec.
Timings for 4608K FFT length (12 cores, 12 workers): 37.38, 37.04, 36.97, 38.33, 38.33, 38.33, 37.33, 37.66, 37.52, 37.28, 36.99, 37.34 ms.  Throughput: 319.71 iter/sec.
Timings for 4608K FFT length (12 cores hyperthreaded, 1 worker):  1.88 ms.  Throughput: 532.43 iter/sec.
Timings for 4608K FFT length (12 cores hyperthreaded, 2 workers):  4.56,  4.56 ms.  Throughput: 438.85 iter/sec.
Timings for 4608K FFT length (12 cores hyperthreaded, 4 workers): 11.56, 11.49, 11.40, 11.44 ms.  Throughput: 348.67 iter/sec.
Timings for 4608K FFT length (12 cores hyperthreaded, 12 workers): 38.53, 37.50, 38.24, 38.10, 38.10, 38.26, 37.91, 37.90, 37.90, 37.46, 38.43, 37.48 ms.  Throughput: 315.95 iter/sec.
Timings for 4800K FFT length (12 cores, 1 worker):  2.00 ms.  Throughput: 498.87 iter/sec.
Timings for 4800K FFT length (12 cores, 2 workers):  4.69,  4.56 ms.  Throughput: 432.88 iter/sec.
Timings for 4800K FFT length (12 cores, 4 workers): 11.97, 11.97, 11.90, 11.91 ms.  Throughput: 335.06 iter/sec.
Timings for 4800K FFT length (12 cores, 12 workers): 39.27, 39.43, 39.40, 39.78, 39.78, 39.76, 39.31, 38.99, 39.31, 38.61, 38.16, 38.82 ms.  Throughput: 306.02 iter/sec.
Timings for 4800K FFT length (12 cores hyperthreaded, 1 worker):  1.91 ms.  Throughput: 524.38 iter/sec.
Timings for 4800K FFT length (12 cores hyperthreaded, 2 workers):  4.87,  4.80 ms.  Throughput: 413.61 iter/sec.
Timings for 4800K FFT length (12 cores hyperthreaded, 4 workers): 12.18, 12.11, 12.05, 12.06 ms.  Throughput: 330.62 iter/sec.
Timings for 4800K FFT length (12 cores hyperthreaded, 12 workers): 40.80, 39.15, 39.41, 39.77, 39.70, 39.57, 38.83, 39.96, 39.93, 39.15, 40.32, 39.10 ms.  Throughput: 302.78 iter/sec.
Timings for 5120K FFT length (12 cores, 1 worker):  2.08 ms.  Throughput: 480.52 iter/sec.
Timings for 5120K FFT length (12 cores, 2 workers):  5.28,  5.13 ms.  Throughput: 384.03 iter/sec.
Timings for 5120K FFT length (12 cores, 4 workers): 12.92, 12.90, 12.88, 12.88 ms.  Throughput: 310.23 iter/sec.
Timings for 5120K FFT length (12 cores, 12 workers): 41.78, 41.79, 41.60, 42.38, 42.38, 42.38, 41.53, 41.42, 41.58, 41.72, 41.72, 41.94 ms.  Throughput: 286.75 iter/sec.
Timings for 5120K FFT length (12 cores hyperthreaded, 1 worker):  2.04 ms.  Throughput: 489.64 iter/sec.
Timings for 5120K FFT length (12 cores hyperthreaded, 2 workers):  5.45,  5.46 ms.  Throughput: 366.66 iter/sec.
Timings for 5120K FFT length (12 cores hyperthreaded, 4 workers): 13.20, 13.09, 13.06, 13.10 ms.  Throughput: 305.06 iter/sec.
[Wed May 06 13:02:31 2020]
Timings for 5120K FFT length (12 cores hyperthreaded, 12 workers): 43.64, 41.83, 42.10, 42.33, 42.65, 42.34, 41.49, 41.64, 43.45, 41.61, 42.61, 42.23 ms.  Throughput: 283.58 iter/sec.
Timings for 5376K FFT length (12 cores, 1 worker):  2.25 ms.  Throughput: 445.33 iter/sec.
Timings for 5376K FFT length (12 cores, 2 workers):  5.68,  5.74 ms.  Throughput: 350.33 iter/sec.
Timings for 5376K FFT length (12 cores, 4 workers): 13.74, 13.54, 13.71, 13.71 ms.  Throughput: 292.51 iter/sec.
Timings for 5376K FFT length (12 cores, 12 workers): 43.53, 43.64, 43.33, 44.65, 44.65, 44.65, 44.59, 44.59, 44.59, 42.48, 42.78, 43.13 ms.  Throughput: 273.54 iter/sec.
Timings for 5376K FFT length (12 cores hyperthreaded, 1 worker):  2.15 ms.  Throughput: 464.80 iter/sec.
Timings for 5376K FFT length (12 cores hyperthreaded, 2 workers):  5.97,  5.99 ms.  Throughput: 334.38 iter/sec.
Timings for 5376K FFT length (12 cores hyperthreaded, 4 workers): 13.91, 13.91, 13.84, 13.74 ms.  Throughput: 288.83 iter/sec.
Timings for 5376K FFT length (12 cores hyperthreaded, 12 workers): 44.56, 44.21, 43.98, 44.74, 44.73, 44.06, 43.96, 43.98, 44.52, 43.44, 45.21, 43.88 ms.  Throughput: 271.08 iter/sec.
Timings for 5600K FFT length (12 cores, 1 worker):  2.40 ms.  Throughput: 417.36 iter/sec.
Timings for 5600K FFT length (12 cores, 2 workers):  6.42,  6.42 ms.  Throughput: 311.38 iter/sec.
Timings for 5600K FFT length (12 cores, 4 workers): 14.66, 14.71, 14.56, 14.55 ms.  Throughput: 273.63 iter/sec.
Timings for 5600K FFT length (12 cores, 12 workers): 46.40, 46.41, 46.22, 46.61, 46.61, 46.60, 45.98, 46.17, 45.71, 45.88, 45.89, 46.73 ms.  Throughput: 259.37 iter/sec.
Timings for 5600K FFT length (12 cores hyperthreaded, 1 worker):  2.24 ms.  Throughput: 445.76 iter/sec.
Timings for 5600K FFT length (12 cores hyperthreaded, 2 workers):  6.57,  6.65 ms.  Throughput: 302.42 iter/sec.
Timings for 5600K FFT length (12 cores hyperthreaded, 4 workers): 14.84, 14.74, 14.72, 14.68 ms.  Throughput: 271.29 iter/sec.
Timings for 5600K FFT length (12 cores hyperthreaded, 12 workers): 48.06, 47.00, 46.31, 46.51, 47.79, 46.89, 46.65, 46.42, 46.93, 46.40, 47.72, 45.97 ms.  Throughput: 255.98 iter/sec.
Timings for 5760K FFT length (12 cores, 1 worker):  2.45 ms.  Throughput: 408.84 iter/sec.
Timings for 5760K FFT length (12 cores, 2 workers):  7.39,  7.19 ms.  Throughput: 274.50 iter/sec.
Timings for 5760K FFT length (12 cores, 4 workers): 15.52, 15.77, 15.32, 15.32 ms.  Throughput: 258.40 iter/sec.
Timings for 5760K FFT length (12 cores, 12 workers): 48.46, 48.20, 48.49, 48.73, 48.30, 48.29, 48.19, 47.87, 48.10, 47.81, 47.50, 48.74 ms.  Throughput: 248.85 iter/sec.
Timings for 5760K FFT length (12 cores hyperthreaded, 1 worker):  2.31 ms.  Throughput: 433.15 iter/sec.
Timings for 5760K FFT length (12 cores hyperthreaded, 2 workers):  7.25,  7.25 ms.  Throughput: 275.81 iter/sec.
Timings for 5760K FFT length (12 cores hyperthreaded, 4 workers): 15.74, 15.62, 15.61, 15.55 ms.  Throughput: 255.93 iter/sec.
Timings for 5760K FFT length (12 cores hyperthreaded, 12 workers): 49.68, 48.17, 49.25, 49.10, 49.21, 48.58, 48.75, 48.99, 48.41, 47.98, 50.80, 47.45 ms.  Throughput: 245.66 iter/sec.
Timings for 6144K FFT length (12 cores, 1 worker):  2.47 ms.  Throughput: 405.59 iter/sec.
Timings for 6144K FFT length (12 cores, 2 workers):  7.15,  7.15 ms.  Throughput: 279.83 iter/sec.
Timings for 6144K FFT length (12 cores, 4 workers): 16.05, 16.02, 15.77, 16.01 ms.  Throughput: 250.63 iter/sec.
[Wed May 06 13:07:41 2020]
Timings for 6144K FFT length (12 cores, 12 workers): 50.89, 51.17, 50.86, 50.14, 49.71, 49.88, 50.37, 50.14, 50.33, 49.77, 49.65, 50.29 ms.  Throughput: 238.75 iter/sec.
Timings for 6144K FFT length (12 cores hyperthreaded, 1 worker):  2.34 ms.  Throughput: 428.14 iter/sec.
Timings for 6144K FFT length (12 cores hyperthreaded, 2 workers):  7.33,  7.33 ms.  Throughput: 272.86 iter/sec.
Timings for 6144K FFT length (12 cores hyperthreaded, 4 workers): 16.20, 16.15, 16.01, 16.11 ms.  Throughput: 248.16 iter/sec.
Timings for 6144K FFT length (12 cores hyperthreaded, 12 workers): 51.51, 51.86, 50.20, 50.99, 51.30, 50.88, 50.22, 50.82, 50.90, 50.07, 52.08, 50.09 ms.  Throughput: 235.75 iter/sec.
Timings for 6400K FFT length (12 cores, 1 worker):  2.64 ms.  Throughput: 378.47 iter/sec.
Timings for 6400K FFT length (12 cores, 2 workers):  7.51,  7.52 ms.  Throughput: 266.15 iter/sec.
Timings for 6400K FFT length (12 cores, 4 workers): 16.70, 16.69, 16.69, 16.69 ms.  Throughput: 239.63 iter/sec.
Timings for 6400K FFT length (12 cores, 12 workers): 52.02, 52.38, 52.32, 53.39, 53.38, 53.41, 52.17, 51.87, 52.04, 52.23, 52.27, 53.04 ms.  Throughput: 228.41 iter/sec.
Timings for 6400K FFT length (12 cores hyperthreaded, 1 worker):  2.44 ms.  Throughput: 410.38 iter/sec.
Timings for 6400K FFT length (12 cores hyperthreaded, 2 workers):  7.76,  7.76 ms.  Throughput: 257.80 iter/sec.
Timings for 6400K FFT length (12 cores hyperthreaded, 4 workers): 17.00, 16.83, 16.83, 16.83 ms.  Throughput: 237.03 iter/sec.
Timings for 6400K FFT length (12 cores hyperthreaded, 12 workers): 53.98, 53.90, 52.14, 53.08, 53.53, 53.08, 53.36, 52.50, 53.01, 53.68, 53.35, 51.77 ms.  Throughput: 225.96 iter/sec.
Timings for 6720K FFT length (12 cores, 1 worker):  2.80 ms.  Throughput: 357.72 iter/sec.
Timings for 6720K FFT length (12 cores, 2 workers):  8.32,  8.23 ms.  Throughput: 241.70 iter/sec.
Timings for 6720K FFT length (12 cores, 4 workers): 18.00, 17.94, 17.87, 17.88 ms.  Throughput: 223.19 iter/sec.
Timings for 6720K FFT length (12 cores, 12 workers): 56.37, 56.39, 55.65, 56.36, 56.29, 55.97, 57.02, 55.47, 55.21, 55.76, 55.47, 55.84 ms.  Throughput: 214.37 iter/sec.
Timings for 6720K FFT length (12 cores hyperthreaded, 1 worker):  2.80 ms.  Throughput: 356.56 iter/sec.
Timings for 6720K FFT length (12 cores hyperthreaded, 2 workers):  8.48,  8.49 ms.  Throughput: 235.73 iter/sec.
Timings for 6720K FFT length (12 cores hyperthreaded, 4 workers): 18.29, 18.10, 18.10, 18.10 ms.  Throughput: 220.40 iter/sec.
Timings for 6720K FFT length (12 cores hyperthreaded, 12 workers): 56.89, 56.14, 57.50, 57.17, 57.31, 56.01, 56.38, 55.81, 57.15, 55.57, 58.23, 55.46 ms.  Throughput: 211.92 iter/sec.
Timings for 7168K FFT length (12 cores, 1 worker):  3.06 ms.  Throughput: 327.04 iter/sec.
Timings for 7168K FFT length (12 cores, 2 workers):  8.78,  8.72 ms.  Throughput: 228.54 iter/sec.
Timings for 7168K FFT length (12 cores, 4 workers): 19.01, 19.01, 18.67, 18.97 ms.  Throughput: 211.49 iter/sec.
Timings for 7168K FFT length (12 cores, 12 workers): 58.41, 58.63, 58.25, 58.97, 59.66, 58.96, 58.67, 58.66, 59.01, 58.13, 58.14, 58.87 ms.  Throughput: 204.45 iter/sec.
Timings for 7168K FFT length (12 cores hyperthreaded, 1 worker):  3.13 ms.  Throughput: 319.17 iter/sec.
Timings for 7168K FFT length (12 cores hyperthreaded, 2 workers):  8.90,  8.90 ms.  Throughput: 224.84 iter/sec.
[Wed May 06 13:12:44 2020]
Timings for 7168K FFT length (12 cores hyperthreaded, 4 workers): 19.34, 19.04, 19.05, 19.09 ms.  Throughput: 209.10 iter/sec.
Timings for 7168K FFT length (12 cores hyperthreaded, 12 workers): 60.66, 59.13, 59.13, 59.45, 59.51, 59.09, 58.57, 58.32, 60.32, 58.40, 61.02, 58.40 ms.  Throughput: 202.29 iter/sec.
Timings for 7680K FFT length (12 cores, 1 worker):  3.66 ms.  Throughput: 273.15 iter/sec.
Timings for 7680K FFT length (12 cores, 2 workers):  9.98,  9.99 ms.  Throughput: 200.25 iter/sec.
Timings for 7680K FFT length (12 cores, 4 workers): 20.94, 20.83, 20.66, 20.83 ms.  Throughput: 192.16 iter/sec.
Timings for 7680K FFT length (12 cores, 12 workers): 64.80, 64.12, 64.43, 64.40, 64.16, 64.19, 63.77, 63.32, 64.23, 63.26, 64.16, 64.29 ms.  Throughput: 187.23 iter/sec.
Timings for 7680K FFT length (12 cores hyperthreaded, 1 worker):  4.48 ms.  Throughput: 223.13 iter/sec.
Timings for 7680K FFT length (12 cores hyperthreaded, 2 workers): 10.27, 10.26 ms.  Throughput: 194.83 iter/sec.
Timings for 7680K FFT length (12 cores hyperthreaded, 4 workers): 21.19, 21.23, 21.11, 21.07 ms.  Throughput: 189.10 iter/sec.
Timings for 7680K FFT length (12 cores hyperthreaded, 12 workers): 66.10, 64.22, 65.39, 66.10, 65.34, 63.92, 64.83, 64.29, 64.78, 64.07, 65.92, 64.05 ms.  Throughput: 184.88 iter/sec.
Timings for 8000K FFT length (12 cores, 1 worker):  3.70 ms.  Throughput: 270.56 iter/sec.
Timings for 8000K FFT length (12 cores, 2 workers): 10.09, 10.10 ms.  Throughput: 198.12 iter/sec.
Timings for 8000K FFT length (12 cores, 4 workers): 21.44, 21.34, 21.36, 21.31 ms.  Throughput: 187.24 iter/sec.
Timings for 8000K FFT length (12 cores, 12 workers): 66.26, 66.15, 65.49, 65.79, 66.02, 65.55, 65.76, 66.20, 66.24, 64.97, 65.31, 65.97 ms.  Throughput: 182.35 iter/sec.
Timings for 8000K FFT length (12 cores hyperthreaded, 1 worker):  4.04 ms.  Throughput: 247.31 iter/sec.
Timings for 8000K FFT length (12 cores hyperthreaded, 2 workers): 10.13, 10.13 ms.  Throughput: 197.46 iter/sec.
Timings for 8000K FFT length (12 cores hyperthreaded, 4 workers): 21.61, 21.56, 21.42, 21.40 ms.  Throughput: 186.08 iter/sec.
Timings for 8000K FFT length (12 cores hyperthreaded, 12 workers): 67.24, 67.97, 65.53, 66.12, 67.14, 67.02, 66.44, 65.96, 66.42, 64.90, 68.01, 65.89 ms.  Throughput: 180.34 iter/sec.
Timings for 8064K FFT length (12 cores, 1 worker):  3.92 ms.  Throughput: 255.09 iter/sec.
Timings for 8064K FFT length (12 cores, 2 workers): 10.28, 10.17 ms.  Throughput: 195.64 iter/sec.
Timings for 8064K FFT length (12 cores, 4 workers): 21.62, 21.50, 21.36, 21.41 ms.  Throughput: 186.29 iter/sec.
Timings for 8064K FFT length (12 cores, 12 workers): 65.68, 66.32, 65.71, 67.01, 67.02, 67.01, 66.23, 66.22, 66.70, 65.68, 65.65, 66.98 ms.  Throughput: 180.87 iter/sec.
Timings for 8064K FFT length (12 cores hyperthreaded, 1 worker):  4.17 ms.  Throughput: 239.72 iter/sec.
Timings for 8064K FFT length (12 cores hyperthreaded, 2 workers): 10.42, 10.42 ms.  Throughput: 191.87 iter/sec.
Timings for 8064K FFT length (12 cores hyperthreaded, 4 workers): 21.77, 21.83, 21.67, 21.67 ms.  Throughput: 184.04 iter/sec.
Timings for 8064K FFT length (12 cores hyperthreaded, 12 workers): 68.76, 66.54, 66.57, 66.41, 69.10, 66.51, 66.44, 66.32, 67.09, 66.31, 68.38, 66.38 ms.  Throughput: 178.97 iter/sec.
Timings for 8192K FFT length (12 cores, 1 worker):  4.03 ms.  Throughput: 248.31 iter/sec.
[Wed May 06 13:17:54 2020]
Timings for 8192K FFT length (12 cores, 2 workers): 10.51, 10.51 ms.  Throughput: 190.34 iter/sec.
Timings for 8192K FFT length (12 cores, 4 workers): 21.95, 21.95, 21.88, 21.88 ms.  Throughput: 182.54 iter/sec.
Timings for 8192K FFT length (12 cores, 12 workers): 67.85, 67.89, 67.81, 67.94, 68.43, 67.94, 68.25, 67.90, 67.90, 65.82, 66.32, 66.94 ms.  Throughput: 177.58 iter/sec.
Timings for 8192K FFT length (12 cores hyperthreaded, 1 worker):  4.38 ms.  Throughput: 228.09 iter/sec.
Timings for 8192K FFT length (12 cores hyperthreaded, 2 workers): 10.59, 10.58 ms.  Throughput: 188.92 iter/sec.
Timings for 8192K FFT length (12 cores hyperthreaded, 4 workers): 22.29, 22.27, 22.01, 22.14 ms.  Throughput: 180.38 iter/sec.
Timings for 8192K FFT length (12 cores hyperthreaded, 12 workers): 72.04, 66.97, 66.99, 68.43, 68.63, 67.56, 66.96, 68.01, 68.29, 67.30, 69.70, 67.36 ms.  Throughput: 176.06 iter/sec.

S485122 · 2020-05-06, 23:54

Quote:

Originally Posted by xx005fs

CPU clockspeed locked at 4.2GHz, running DDR4 3733 with Dual channel and Dual Rank memory, FCLK at 1866MHz.

Accoding to that benchmark
For a FFT size of 2048k the best configuration is 4 workers not using hyper-threading.
For FFT sizes from 2240k to 3584k it is 2 workers using hyper-threading.
For FFT sizes from 3840k to 4096k it is 2 workers not using hyper-threading.
For FFT sizes from 4480k to 6400k it is 1 worker using hyper-threading.
For FFT sizes from 6720k to 8192k it is 1 worker not using hyper-threading.

The difference between using HT or not varies between almost 20 % to less than 1 %.
At some FFT sizes the difference is much bigger than for the adjoining sizes, maybe the effect of background tasks ? The best configurations do show a pattern.

Jacob

xx005fs · 2020-05-07, 17:00

Quote:

Originally Posted by S485122

Accoding to that benchmark
For a FFT size of 2048k the best configuration is 4 workers not using hyper-threading.
For FFT sizes from 2240k to 3584k it is 2 workers using hyper-threading.
For FFT sizes from 3840k to 4096k it is 2 workers not using hyper-threading.
For FFT sizes from 4480k to 6400k it is 1 worker using hyper-threading.
For FFT sizes from 6720k to 8192k it is 1 worker not using hyper-threading.

The difference between using HT or not varies between almost 20 % to less than 1 %.
At some FFT sizes the difference is much bigger than for the adjoining sizes, maybe the effect of background tasks ? The best configurations do show a pattern.

Jacob

It seems that Hyperthreading is only significant for smaller FFT size since Ryzen 3000 series has a rather large cache. At the same time on smaller FFTs power consumption also goes through the roof without any power limit in place. I'm running my 3900x on 1.1V at 4.2GHz and for the bigger FFTs it was around 105-115W, while for 2048K it was more like 175W with hyperthreading and 4 workers, which also indicates better utilization of the FPU.

For the big deltas in some FFTs it could be background tasks since I ran this on Windows, but I did make sure to kill as many as possible.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Perpetual "interesting video" thread...	Xyzzy	Lounge	35	2021-02-28 06:57
LLR benchmark thread	Oddball	Riesel Prime Search	5	2010-08-02 00:11
Perpetual I'm pi**ed off thread	rogue	Soap Box	19	2009-10-28 19:17
Perpetual autostereogram thread...	Xyzzy	Lounge	10	2006-09-28 00:36
Perpetual ECM factoring challenge thread...	Xyzzy	Factoring	65	2005-09-05 08:16