Skip to content

Latest commit

 

History

History
91 lines (75 loc) · 1.89 KB

Findings.md

File metadata and controls

91 lines (75 loc) · 1.89 KB

List of any and all performance anomalies in Storage-N systems

Cinebench :

  • Storage-0:

    • GPU : 6958
    • CPU-MultiCore 897
    • CPU-SingleCore 110
  • Storage-1:

    • GPU 6935
    • CPU-MultiCore 918
    • CPU-SingleCore 111
  • Storage-2:

    • GPU 6942
    • CPU-MultiCore 910
    • CPU-SingleCore 111
  • Storage-3:

    • GPU 6920
    • CPU-MultiCore 920
    • CPU-SingleCore 110

No significant differences

Corona

  • Storage-0 : 5,109,480

  • Storage-1 : 5,288,234

  • Storage-2 : 5,296,405

  • Storage-3 : 5,252,086

Storage-0 multi-core was slightly worse, but well within margin of error.

Geekbench

Storage-0

  • Single Core : 2577

  • Multi Core : 15197

  • GPU-OpenCL : 74102

  • GPU-Vulkan : 71016

Storage-1

  • Single Core : 2591

  • Multi Core : 15383

  • GPU-OpenCL : 73667

  • GPU-Vulkan : 70818

Storage-2

  • Single Core : 2576

  • Multi Core : 15194

  • GPU-OpenCL : 75250

  • GPU-Vulkan : 71486

Storage-3

  • Single Core : 2594

  • Multi Core : 15148

  • GPU-OpenCL : 75911

  • GPU-Vulkan : 70200

Storage-0 seems on par here, so everything should be fine. GPUs are in spec as well.

3D-Mark :

  • Storage-0

    • TimeSpy : 6436

    • FireStrike : 15127

  • Storage-1

    • TimeSpy : 6369

    • FireStrike : 15162

  • Storage-2

    • TimeSpy : 6551

    • FireStrike : 15477

  • Storage-3

    • TimeSpy : 6419

    • FireStrike : 15124

Compubench :

OpenCL :

Everything looks fine

Cuda :

Everything looks fine

CrystalDiskMark :

  • SSDs:

    • Random Reads show variations (60-90), everything else is consistent

  • HDDs:

    • Similar random reads and writes, but sequential shows around a 5-10% variation.

    • Some drives show results ranging from 205-210 whereas some are in the low 190s.

S.M.A.R.T :

  • No increase in negative indicators.

Overall Results :

  • Systems are performing more or less within spec.