From: Georgi Gerganov Date: Tue, 30 Sep 2025 10:28:50 +0000 (+0300) Subject: bench : add rtx 5090 [no ci] X-Git-Tag: upstream/1.8.0~3 X-Git-Url: https://git.djapps.eu/?a=commitdiff_plain;h=1e5ad50f8f9cb26539df3c91ee300c5391347613;p=pkg%2Fggml%2Fsources%2Fwhisper.cpp bench : add rtx 5090 [no ci] --- diff --git a/scripts/bench-all-gg.txt b/scripts/bench-all-gg.txt index a09126da..82bf6aa1 100644 --- a/scripts/bench-all-gg.txt +++ b/scripts/bench-all-gg.txt @@ -247,6 +247,43 @@ make -j && ./scripts/bench-all.sh 1 1 1 | M4 Max | METAL | large-v3-turbo | 1 | 1 | 250.19 | 1.64 | 0.37 | 0.04 | a77d11d9 | +# RTX 5090 + +make -j && ./scripts/bench-all.sh 1 1 0 + +| GPU | Config | Model | Th | FA | Enc. | Dec. | Bch5 | PP | Commit | +| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | +| RTX 5090 | CUDA | tiny | 1 | 0 | 2.06 | 0.55 | 0.13 | 0.00 | e4bf87b0 | +| RTX 5090 | CUDA | tiny-q8_0 | 1 | 0 | 2.50 | 0.55 | 0.14 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | base | 1 | 0 | 3.72 | 0.81 | 0.19 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | base-q8_0 | 1 | 0 | 4.35 | 0.79 | 0.20 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | small | 1 | 0 | 11.24 | 1.55 | 0.38 | 0.02 | e4bf87b0 | +| RTX 5090 | CUDA | small-q8_0 | 1 | 0 | 12.69 | 1.69 | 0.40 | 0.02 | e4bf87b0 | +| RTX 5090 | CUDA | medium | 1 | 0 | 31.16 | 3.19 | 0.79 | 0.04 | e4bf87b0 | +| RTX 5090 | CUDA | medium-q8_0 | 1 | 0 | 32.74 | 3.43 | 0.80 | 0.05 | e4bf87b0 | +| RTX 5090 | CUDA | large-v2 | 1 | 0 | 50.09 | 4.55 | 1.14 | 0.05 | e4bf87b0 | +| RTX 5090 | CUDA | large-v2-q8_0 | 1 | 0 | 52.44 | 4.76 | 1.11 | 0.07 | e4bf87b0 | +| RTX 5090 | CUDA | large-v3-turbo | 1 | 0 | 46.78 | 0.70 | 0.17 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | large-v3-turbo-q8_0 | 1 | 0 | 48.57 | 0.70 | 0.16 | 0.01 | e4bf87b0 | + +make -j && ./scripts/bench-all.sh 1 1 1 + +| GPU | Config | Model | Th | FA | Enc. | Dec. | Bch5 | PP | Commit | +| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | +| RTX 5090 | CUDA | tiny | 1 | 1 | 1.39 | 0.47 | 0.11 | 0.00 | e4bf87b0 | +| RTX 5090 | CUDA | tiny-q8_0 | 1 | 1 | 1.83 | 0.48 | 0.12 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | base | 1 | 1 | 2.17 | 0.70 | 0.16 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | base-q8_0 | 1 | 1 | 2.78 | 0.68 | 0.17 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | small | 1 | 1 | 5.02 | 1.33 | 0.32 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | small-q8_0 | 1 | 1 | 6.39 | 1.46 | 0.34 | 0.02 | e4bf87b0 | +| RTX 5090 | CUDA | medium | 1 | 1 | 13.89 | 2.68 | 0.64 | 0.03 | e4bf87b0 | +| RTX 5090 | CUDA | medium-q8_0 | 1 | 1 | 15.40 | 2.92 | 0.67 | 0.04 | e4bf87b0 | +| RTX 5090 | CUDA | large-v2 | 1 | 1 | 21.24 | 3.88 | 0.96 | 0.04 | e4bf87b0 | +| RTX 5090 | CUDA | large-v2-q8_0 | 1 | 1 | 23.54 | 4.01 | 0.93 | 0.05 | e4bf87b0 | +| RTX 5090 | CUDA | large-v3-turbo | 1 | 1 | 18.18 | 0.62 | 0.15 | 0.01 | e4bf87b0 | +| RTX 5090 | CUDA | large-v3-turbo-q8_0 | 1 | 1 | 19.89 | 0.61 | 0.14 | 0.01 | e4bf87b0 | + + # V100 GGML_CUDA=1 make -j && ./scripts/bench-all.sh 8 1 0