Commit 27a31409 authored by muozturk's avatar muozturk
Browse files

Readme

parent 80fc636b
# Instructions for ```example_contraction_bilinear_xdl_fp32``` # Instructions for ```example_complex_contraction_bilinear_xdl_fp32```
## Run ## Run
```bash ```bash
...@@ -8,13 +8,16 @@ ...@@ -8,13 +8,16 @@
./bin/example_contraction_bilinear_xdl_fp32 1 1 1 ./bin/example_contraction_bilinear_xdl_fp32 1 1 1
``` ```
Result (MI100 @ dynammic freq, 46TFlops peak FP32) Result (MI210 )
``` ```
a_ms_ks: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1} a_ms_ks_re: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
b_ks_ns: dim 4, lengths {32, 64, 32, 64}, strides {128, 1, 524288, 4096} b_ns_ks_re: dim 4, lengths {32, 64, 32, 64}, strides {524288, 4096, 128, 1}
c_ms_ns: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1} d_ms_ns_re: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
launch_and_time_kernel: grid_dim {240, 1, 1}, block_dim {256, 1, 1} e_ms_ns_re: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
Warm up 1 time a_ms_ks_img: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
Start running 10 times... b_ns_ks_img: dim 4, lengths {32, 64, 32, 64}, strides {524288, 4096, 128, 1}
Perf: 0.843286 ms, 38.1985 TFlops, 94.5014 GB/s, DeviceContractionMultipleD_Xdl_CShuffle<256, 256, 128, 16, 4, 4> d_ms_ns_img: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
e_ms_ns_img: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
Perf: 4.51253 ms, 14.2768 TFlops, 31.6023 GB/s,
DeviceContractionMultipleD_Xdl_CShuffle<2, 2, 2, 256, 256, 128, 16, 4, 4, 2, 2>
``` ```
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment