Just some test scores questions.

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • rbtree11
    Junior Member
    • Nov 2020
    • 13

    #1

    Just some test scores questions.

    I've recently built my latest Ryzen PC : 9950x, 2x48 gb CL6400 DDR5 Ram, x870E ASRock Taichi Motherboard.

    The first file is the PC, basically stock. I think PBO was set in the BIOS. The CPU score is about 1.5% above baseline/average . So, I got good silicone!! Memory was set to XMP. Quite sure it was at its rated 6400 CL 32. I ran several 3D tests with the EVGA 3090 Ultra at varying overclock settings. The scores reached almost 900 points above that 33343 mark. (Way above your charts)

    Next files show a very good CB 23 score (almost as good as an overclocked 9950X3D score in your chart- none showed for an o/c 9950x) after doing some basic overclocking in the bios and setting the memory to "extreme" which bumped the memory way above where it had been, which was close to my DDR4 3600 ram scores of 3600-3700. (I've never overclocked memory as I don't understand how, I just choose the XMP settings.

    So, from the first file to the next two, there is one other change, which I wonder if that is why the Disk score dropped way off. I added 3 more NVMe drives, all fast Gen 4 as the chipset/mbonly has one PCI 5 m.2slot. I'd like to know why the disk scores dropped so much, seeing as PM supposedly only checks the C drive. Also, I checked the drive in Crystal Diskmark and it was where it usually is, about 13000/11000. That's one head scratcher.

    The other head scratcher is the now lower 3D scores--I tested the GPU at stock and two different overclock settings, which worked before, but this time, all 3 scores were close to 30000, 3-4k lower. I incresedd the fan speeds to keep the temps down, but that had no effect.

    I understand that there can be variances from test to test... but don't understand why my CPU/ram bios setting changes should affect the 3D scores. (2D was about the same.)

    Sure, it's really just for fun, but I was hoping to really get a great score, but the two strange issues kept that from happening.

    Thanks in advance! For as much as I enjoy using PM. what with having at least one too many PC's (4) and two laptops that work, I should break down and pay to join in the fun, and be legit!

    Roger
    Attached Files
  • David (PassMark)
    Administrator
    • Jan 2003
    • 11099

    #2
    Disk scores can be highly variable. Even a small amount of background activity can have a big effect. As can the duration of the test and how much sustained load there has been on the drive.
    So you probably need to run the test several times per config setup, with a significant pause between each test run to allow for cooling and caching effects to disappear.

    We can't really comment on the Cinebench stuff. It isn't our software.

    Would have been interesting to see both the before and after results for the individual 3D tests. But this could be be background activity or temperature related issues.

    Comment

    Working...