... | ... | @@ -11,22 +11,11 @@ The following benchmarks indicate how the Librem 5 performs compared to other ha |
|
|
| ↓ FPU Raytracing | 1.96 | 2.99 | 4.81 | 7.12 | -61% | -138% | 48% |
|
|
|
|------------------|--------|--------|--------|--------|-------|-------|------|
|
|
|
| ↑ glmark2-es2-wayland| 1967 | 267† | 203 | 84 | -32%† | -218%†| 142% |
|
|
|
| - build | 2493 | 336 | 467 | 214 | 28% | -57% | 119% |
|
|
|
| - shading | 2090 | 294 | 206 | 75 | -43% | -294% | 176% |
|
|
|
| - bump | 2110 | 298 | 188 | 57 | -59% | -423% | 230% |
|
|
|
| - effect2d | 1820 | 169 | 91 | 27 | -87% | -538% | 242% |
|
|
|
| - pulsar | 3168 | 326 | 336 | 210 | 3% | -55% | 60% |
|
|
|
| - buffer | 541 | 159 | 129 | 63 | -23% | -153% | 105% |
|
|
|
| - ideas | 1519 | 291 | 202 | 59 | -44% | -393% | 242% |
|
|
|
| - jellyfish | 2507 | 200 | 79 | 19 | -153% | -953% | 316% |
|
|
|
| - conditionals | 2604 | 317 | 239 | 87 | -33% | -265% | 175% |
|
|
|
| - function | 2609 | 286 | 124 | 50 | -131% | -471% | 147% |
|
|
|
| - loop | 2463 | 261 | 123 | 52 | -112% | -402% | 137% |
|
|
|
|
|
|
↑ Benchmark where higher is better.
|
|
|
↓ Benchmark where lower is better.
|
|
|
\* glmark2-es2-wayland was compiled without the “terrain” scene (which isn’t able to run on the PinePhone).
|
|
|
† Due to a [bug](https://github.com/glmark2/glmark2/issues/124) with GLMark2 in the Raspberry Pi, it was compiled without the "texture", "windows", "terrain", "shadow" and "refract" scenes, so the Raspberry Pi score should be treated with caution when comparing to other devices.
|
|
|
\* glmark2-es2-wayland was compiled without the “terrain” scene (which isn’t able to run on the PinePhone).
|
|
|
† Due to a [bug](https://github.com/glmark2/glmark2/issues/124) in GLMark2 with the Raspberry Pi 4B, it was compiled without the "texture", "windows", "terrain", "shadow" and "refract" scenes, so the Raspberry Pi score should be treated with caution when comparing to other devices.
|
|
|
**Sources:** [hardinfo](https://github.com/lpereira/hardinfo/) and [glmark2](https://github.com/glmark2/glmark2) tests by Sebastian Krzyszkowiak (https://dosowisko.net/bench/) and Amos Batto (https://github.com/amosbatto/benchmarks). See the [forum discussion](https://forums.puri.sm/t/still-no-dogwood/10034/55) about the tests.
|
|
|
|
|
|
**Information about the benchmarks:**
|
... | ... | |