MPI v1.0
The Maag Performance Index
One number for the whole system, not just the graphics card. My baseline rig scores 100. Everything else is measured against it at 2560x1440, on the settings I actually play with.
Leaderboard
| # | System | Gaming /55 | 3DMark /15 | CPU /25 | Storage /5 | MPI |
|---|---|---|---|---|---|---|
| 1 | 13900K + RTX 5090BaselineIntel Core i9-13900K · ASUS ROG Astral RTX 5090 OC · 32GB DDR5-7200 CL34 | 55.00 | 15.00 | 25.00 | 5.00 | 100.00 |
| 9850X3D + RTX 5090AMD Ryzen 7 9850X3D · ASUS ROG Astral RTX 5090 OC · 48GB DDR5-8000 CL38Score drops with the video | ||||||
| 270K Plus + RTX 5070 TiIntel Core Ultra 7 270K Plus · Zotac RTX 5070 Ti Solid SFF OC · 32GB DDR5-7200 CL34Not tested yet | ||||||
Scores aren't capped: a rig can beat the baseline and land above 100.
How the score is built
100 points across four areas. A rig earns points in each test by how it compares to the baseline.
Gaming
55 ptsFive games, 11 points each. Average FPS counts for 70% and 1% lows for 30%. No frame generation.
- Cyberpunk 2077
- Gears of War: E-Day
- Forza Horizon 6
- Total War: Warhammer III
- Modern Warfare 4
3DMark
15 ptsThree tests, 5 points each. Graphics score only, so the CPU isn't counted twice.
- Steel Nomad
- Speed Way
- Time Spy Extreme
CPU and productivity
25 ptsFive tests, 5 points each.
- Cinebench R23 single-core
- Cinebench R23 multi-core
- Blender Benchmark (CPU)
- 7-Zip Benchmark
- Adobe Premiere export time
Storage
5 ptsOne test on the system drive.
- UL Full System Drive Benchmark
The rules
- Every result is the median of three runs.
- A crash anywhere means the whole suite gets run again.
- Standard 2560x1440, not ultrawide, so midrange cards get a fair shot.
- Daily-driver settings: the tune I actually use, not a one-off benchmark tune.
- The baseline rig is locked for all of v1.0. It never gets re-tested to move the goalposts.