Announcement

Collapse
No announcement yet.

Phoronix Test Suite 9.4 M3 Released With More Improvements For Benchmark Result Analysis

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Phoronix Test Suite 9.4 M3 Released With More Improvements For Benchmark Result Analysis

    Phoronix: Phoronix Test Suite 9.4 M3 Released With More Improvements For Benchmark Result Analysis

    The third and likely last test release of Phoronix Test Suite 9.4-Vestby is now available for your cross-platform, open-source benchmarking needs...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    As always, Michael you rock, and thank you for PTS - sounds like another exciting release.

    I was just running some PTS this morning (I was curious about PGO and stumbled across an old phoronix post).

    I haven't quite figured out the PGO testing but I was able to conduct a C-Ray test with three different set of CFLAGS/CXXFLAGS set (no flags, -O3 -march=native, and ClearLinux flags).

    45 second difference (!!) just by adding -O3 -march=native. The aggressive ClearLinux flags improved it only slightly (for this particular test at least).

    Code:
    [B]phoronix-test-suite - master ❯ CFLAGS="" CXXFLAGS="" ./phoronix-test-suite pgo.benchmark c-ray[/B]
    C-Ray 1.1:
        pts/c-ray-1.2.0
        Test 1 of 1
        Estimated Trial Run Count:    3                    
        Estimated Time To Completion: 8 Minutes [09:13 PST]
            Started Run 1 @ 09:06:45
            Started Run 2 @ 09:09:51
            Started Run 3 @ 09:12:57
    
    [B]   Total Time - 4K, 16 Rays Per Pixel:
            181.701
            181.712
            181.79
    
        Average: 181.734 Seconds[/B]
    
    ----
    
    [B]phoronix-test-suite - master ❯ CFLAGS="-O3 -march=native" CXXFLAGS="-O3 -march=native" ./phoronix-test-suite pgo.benchmark c-ray[/B]
    C-Ray 1.1:
        pts/c-ray-1.2.0
        Test 1 of 1
        Estimated Trial Run Count:    3                    
        Estimated Time To Completion: 9 Minutes [09:28 PST]
            Started Run 1 @ 09:20:01
            Started Run 2 @ 09:22:23
            Started Run 3 @ 09:24:44
    
    [B]   Total Time - 4K, 16 Rays Per Pixel:
            137.049
            137.294
            137.195
    
        Average: 137.179 Seconds[/B]
    
    -----
    
    ~ ❯ echo $CFLAGS
    -g -O3 -feliminate-unused-debug-types -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=32 -Wformat -Wformat-security -m64 -fasynchronous-unwind-tables -Wp,-D_REENTRANT -ftree-loop-distribute-patterns -Wl,-z -Wl,now -Wl,-z -Wl,relro -fno-semantic-interposition -ffat-lto-objects -fno-trapping-math -Wl,-sort-common -Wl,--enable-new-dtags -march=haswell -mtune=haswell
    
    ~ ❯ echo $CXXFLAGS
    -g -O3 -feliminate-unused-debug-types -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=32 -Wformat -Wformat-security -m64 -fasynchronous-unwind-tables -Wp,-D_REENTRANT -ftree-loop-distribute-patterns -Wl,-z -Wl,now -Wl,-z -Wl,relro -fno-semantic-interposition -ffat-lto-objects -fno-trapping-math -Wl,-sort-common -Wl,--enable-new-dtags -march=haswell -mtune=haswell -fvisibility-inlines-hidden -Wl,--enable-new-dtags
    
    [B]phoronix-test-suite - master ❯ ./phoronix-test-suite pgo.benchmark c-ray[/B]
    
    C-Ray 1.1:
        pts/c-ray-1.2.0
        Test 1 of 1
        Estimated Trial Run Count:    3                      
        Estimated Time To Completion: 16 Minutes [09:08 PST]
            Started Run 1 @ 08:53:35
            Started Run 2 @ 08:55:56
            Started Run 3 @ 08:58:17
    
    [B]Total Time - 4K, 16 Rays Per Pixel:
            136.831
            136.602
            136.678[/B]
    
    [B]Average: 136.704 Seconds[/B]

    Comment


    • #3
      I am waiting for the change of the Steam default folder to appear. Or are there already workarounds?

      Comment


      • #4
        Originally posted by guglovich View Post
        I am waiting for the change of the Steam default folder to appear. Or are there already workarounds?
        What change? Currently all PTS test profiles rely upon the default Steam location. Happy to review any patches but no other changes are planned on my side in just doing the default and already being very quirky having to script around Steam and not having any enterprise customers I am aware of using the Steam tests.
        Michael Larabel
        https://www.michaellarabel.com/

        Comment


        • #5
          I use steam through flatpak, anyway to make that work? Symlinks, or changing the way it's called? Thanks.

          Comment


          • #6
            Looking ahead to 9.5 -

            On my system, the disk benchmarking tests generally fail to run at all. I'll SWAG that they attempt to do low-level "disk" I/O on the device, but the device which I would like to benchmark is an NVMe SSD (PCIe 4x4). It would be nice to offer some "disk" I/O benchmarks which use higher-level disk I/O requests, avoiding incompatibility with NVMe devices.

            Comment


            • #7
              Originally posted by Michael View Post

              What change? Currently all PTS test profiles rely upon the default Steam location. Happy to review any patches but no other changes are planned on my side in just doing the default and already being very quirky having to script around Steam and not having any enterprise customers I am aware of using the Steam tests.

              But this is often inconvenient. For example, delete games with SSD to test another game, delete, and reinstall my game. Another thing to install on the HDD and specify the path in Phoronix

              Comment

              Working...
              X