HWBOT 3DMark03 Benchmark Validation Regulations

  • Benchmark Rules
  • 2
HWBOT 3DMark03 Benchmark Validation Regulations – HWBOT Content Site

HWBOT 3DMark03 Benchmark Validation Regulations

Allowed benchmark versions
Allowed operating systems
Default benchmark settings1024×768, No AA, Optimal Texture Filtering, Max Aniso: 4, Optimal Vertex Shaders.
ScoreThe score is obtained after running the four game tests.
Allowed optimisations
  • Disable unnecessary tests
  • LOD
Disallowed tweaks/cheats
  • Altering benchmark files or the rendering
  • Any software or human interaction altering the perceived speed of the benchmark program, tricking it to believe it ran faster
  • Lucid Virtu MVP
  • Mipmap
  • Change benchmark settings
Verification screenshot requirements
  • CPU-Z CPU tab (version 1.91 or newer)
  • CPU-Z Memory tab (version 1.91 or newer)
  • GPU-Z Graphics Card tab
  • Benchmark window
  • Benchmark score window
  • Benchmark subdetails window
  • Full screenshot required, no clipping
Verification linkVerification link required for Global Top 20
Example screenshot

Red: Mandatory Green: Optional

Additional notesThis page contains the benchmark result regulation specifications for 3DMark03. The rules as described in our General Rules section are still applicable even if they are not specifically mentioned on this page.

It is your responsibility, and the responsibility of your team captain(s) that these rules are followed properly. Your submission may be blocked by HWBOT moderators without notice if your submission is in violation of one of these rules. Multiple infractions may lead to a warning, and eventually an account ban.

  • Feb 01, 2020: updated CPUZ version and full screenshot requirementFeb 15, 2017: link
  • May 15, 2008: Initial draw

Join the Conversation

No comments

Leave a comment

Christian Ney says:








Belgium leeghoofd says:

New 2020 scores: Require CPU-Z 1.91 or newer version

New 2020 scores: Full desktop screenshot required ( no clipping)

Please log in or register to comment.