Skip to content
KernelIndex
Search⌘K
Sign in

submission 32127

KernelBot user 213 · python · License unknown
Kernel source · 1 lines ↓

Use it

Vendorable · source mirrored · license unknownView source →

No package. Vendor the mirrored source: 1 lines, June 9 Researcher Reciprocity License v1.0.

submission.py
curl "https://kernelindex.com/api/v1/implementations/kernelbot-amd-fp8-mm-32127?include=source"
interfacepython
Compatibility
measured onAMD Instinct MI300X
declared hardwareAMD Instinct MI300X
architecturesgfx942
dtypesfp32, fp8_e4m3fnuz

Benchmark evidence

18 measurements across 1 GPU, fastest first.

Operation / workload
Hardware
Latency
Rank
Observed
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 512]
AMD Instinct MI300X
45.2µs
#47 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 256]
AMD Instinct MI300X
45.4µs
#43 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 1536]
AMD Instinct MI300X
59.6µs
#36 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=512 · seed=6563
AMD Instinct MI300X
71.4µs
#23 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=576 · seed=12346
AMD Instinct MI300X
72.4µs
#23 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=1536 · seed=8135
AMD Instinct MI300X
108.6µs
#21 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 2048]
AMD Instinct MI300X
124.3µs
#41 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 512]
AMD Instinct MI300X
134.7µs
#58 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 2304]
AMD Instinct MI300X
136.4µs
#40 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 256]
AMD Instinct MI300X
141.4µs
#69 of 98
2025-06-15
Run →
Show all 18 measurements ›
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 1536]
AMD Instinct MI300X
193.0µs
#44 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=512 · seed=12341
AMD Instinct MI300X
198.0µs
#43 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=576 · seed=9863
AMD Instinct MI300X
206.2µs
#40 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=4608 · seed=7531
AMD Instinct MI300X
243.1µs
#50 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=1536 · seed=6543
AMD Instinct MI300X
401.7µs
#48 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 2048]
AMD Instinct MI300X
492.4µs
#47 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 2304]
AMD Instinct MI300X
543.7µs
#45 of 98
2025-06-15
Run →
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=4608 · seed=65436
AMD Instinct MI300X
1.09ms
#46 of 98
2025-06-15
Run →

Reported · How evidence levels are derived →

Source and license

sourceavailable
revision digestsha256:e01185e04ceeac114eb02b2017a6606b14ecc4c59dc21ab7f58f1ba33266fb81
license declaredunknown
license concludedunknown
imported2026-08-15

Kernel source

submission.py1 lines
IyBUaGlzIHNjcmlwdCBwcm92aWRlcyBhIHRlbXBsYXRlIGZvciB1c2luZyBsb2FkX2lubGluZSB0byBydW4gYSBISVAga2VybmVsIGZvcgpmcm9tIHRvcmNoLnV0aWxzLmNwcF9leHRlbnNpb24gaW1wb3J0IGxvYWRfaW5saW5lCmZyb20gdGFzayBpbXBvcnQgaW5wdXRfdCwgb3V0cHV0X3QKQ1BQX1dSQVBQRVIgPSAiIiIKdm9pZCBmcDhfbW0odG9yY2g6OlRlbnNvciBhLCB0b3JjaDo6VGVuc29yIGIsIHRvcmNoOjpUZW5zb3IgYXMsIHRvcmNoOjpUZW5zb3IgYnMsIHRvcmNoOjpUZW5zb3IgYyk7CiIiIgoKQ1VEQV9TUkMgPSAiIiIKI2luY2x1ZGUgPGhpcC9hbWRfZGV0YWlsL2FtZF9oaXBfZnA4Lmg+CiNpbmNsdWRlIDxoaXAvYW1kX2RldGFpbC9hbWRfaGlwX2JmMTYuaD4KCnVzaW5nIGYzMngxNiA9IF9fYXR0cmlidXRlX18oIChfX3ZlY3Rvcl9zaXplX18oMTYgKiBzaXplb2YoZmxvYXQpKSApKSBmbG9hdDsKCmNvbnN0ZXhwciBjb25zdCBzaXplX3QgQkxPQ0sgPSAxMjg7CgpfX2RldmljZV9fIHN0YXRpYyB1aW50NjRfdCBwYWNrX3UxNng4X3RvX3U2NChjb25zdCB1aW50MTZfdCogc3JjKQp7CiAgICB1aW50MTZfdCBkc3RbXSA9IHsgc3JjWzFdLCBzcmNbM10sIHNyY1s1XSwgc3JjWzddIH07CiAgICByZXR1cm4gKnJlaW50ZXJwcmV0X2Nhc3Q8dWludDY0X3QqPihkc3QpOwp9CgpfX2dsb2JhbF9fIHZvaWQgY3VzdG9tX2tlcm5lbChjb25zdCB1aW50NjRfdCogY2FjaGUsIGNvbnN0IGZsb2F0KiBhcywgY29uc3QgZmxvYXQqIGJzLCAKICAgICAgICAgICAgICAgICAgIF9faGlwX2JmbG9hdDE2KiBjLCBzaXplX3QgbSwgc2l6ZV90IG4sIHNpemVfdCBrLCBzaXplX3QgY29sLCBzaXplX3Qgb2Zmc2V0KSB7CiAgICBjb25zdCB1aW50NjRfdCogYSA9IGNhY2hlOwogICAgY29uc3QgdWludDY0X3QqIGIgPSBjYWNoZSArIG9mZnNldDsKICAgIC8vIDMyeDMyIHRpbGVzIHdpdGggc2luZ2xlIHdhdmVmcm9udCAoNjQgdGhyZWFkcykKICAgIHNpemVfdCBieCA9IGJsb2NrSWR4Lng7CiAgICBzaXplX3QgYnkgPSBibG9ja0lkeC55OwogICAgc2l6ZV90IGN4ID0gMzIgKiBieDsKICAgIHNpemVfdCBjeSA9IDMyICogYnk7CiAgICBpZihjeCA+PSBtIHx8IGN5ID49IG4pIHJldHVybjsKICAgIHNpemVfdCB0eCA9IHRocmVhZElkeC54ICYgMzE7CiAgICBzaXplX3QgdHkgPSB0aHJlYWRJZHgueCA+PiA1OwogICAgc2l6ZV90IHNuID0gKG4gKyBCTE9DSyAtIDEpID4+IDc7CiAgICBicyArPSBjeSA+PiA3OwogICAgYXMgKz0gY3ggKyB0eDsKICAgIGEgKz0gYnggKiBjb2wgKyB0aHJlYWRJZHgueDsKICAgIGIgKz0gYnkgKiBjb2wgKyB0aHJlYWRJZHgueDsKICAgIGYzMngxNiByZXN1bHQgPSB7IH07CiAgICBmb3Ioc2l6ZV90IGkgPSAwOyBpIDwgazsgaSArPSBCTE9DSykgewogICAgICAgIGNvbnN0IGF1dG8qIGFwID0gYSArIGkgKiA0OwogICAgICAgIGNvbnN0IGF1dG8qIGJwID0gYiArIGkgKiA0OwogICAgICAgIGYzMngxNiBibG9ja19yZXN1bHQgPSB7IH07CiAgICAgICAgCiAgICAgICAgYmxvY2tfcmVzdWx0ID0gX19idWlsdGluX2FtZGdjbl9tZm1hX2YzMl8zMngzMngxNl9mcDhfZnA4KGJwWzBdLCBhcFswXSwgYmxvY2tfcmVzdWx0LCAwLCAwLCAwKTsKICAgICAgICBibG9ja19yZXN1bHQgPSBfX2J1aWx0aW5fYW1kZ2NuX21mbWFfZjMyXzMyeDMyeDE2X2ZwOF9mcDgoYnBbMHg0MF0sIGFwWzB4NDBdLCBibG9ja19yZXN1bHQsIDAsIDAsIDApOwogICAgICAgIGJsb2NrX3Jlc3VsdCA9IF9fYnVpbHRpbl9hbWRnY25fbWZtYV9mMzJfMzJ4MzJ4MTZfZnA4X2ZwOChicFsweDgwXSwgYXBbMHg4MF0sIGJsb2NrX3Jlc3VsdCwgMCwgMCwgMCk7CiAgICAgICAgYmxvY2tfcmVzdWx0ID0gX19idWlsdGluX2FtZGdjbl9tZm1hX2YzMl8zMngzMngxNl9mcDhfZnA4KGJwWzB4YzBdLCBhcFsweGMwXSwgYmxvY2tfcmVzdWx0LCAwLCAwLCAwKTsKICAgICAgICBibG9ja19yZXN1bHQgPSBfX2J1aWx0aW5fYW1kZ2NuX21mbWFfZjMyXzMyeDMyeDE2X2ZwOF9mcDgoYnBbMHgxMDBdLCBhcFsweDEwMF0sIGJsb2NrX3Jlc3VsdCwgMCwgMCwgMCk7CiAgICAgICAgYmxvY2tfcmVzdWx0ID0gX19idWlsdGluX2FtZGdjbl9tZm1hX2YzMl8zMngzMngxNl9mcDhfZnA4KGJwWzB4MTQwXSwgYXBbMHgxNDBdLCBibG9ja19yZXN1bHQsIDAsIDAsIDApOwogICAgICAgIGJsb2NrX3Jlc3VsdCA9IF9fYnVpbHRpbl9hbWRnY25fbWZtYV9mMzJfMzJ4MzJ4MTZfZnA4X2ZwOChicFsweDE4MF0sIGFwWzB4MTgwXSwgYmxvY2tfcmVzdWx0LCAwLCAwLCAwKTsKICAgICAgICBibG9ja19yZXN1bHQgPSBfX2J1aWx0aW5fYW1kZ2NuX21mbWFfZjMyXzMyeDMyeDE2X2ZwOF9mcDgoYnBbMHgxYzBdLCBhcFsweDFjMF0sIGJsb2NrX3Jlc3VsdCwgMCwgMCwgMCk7CiAgICAgICAgCiAgICAgICAgcmVzdWx0ICs9IGJsb2NrX3Jlc3VsdCAqIGJzWyhpID4+IDcpICogc25dICogYXNbKGkgPj4gNykgKiBtXTsKICAgIH0KICAgIGNvbnN0IGF1dG8qIHNyYyA9IHJlaW50ZXJwcmV0X2Nhc3Q8Y29uc3QgdWludDE2X3QoKilbOF0+KCZyZXN1bHQpOwogICAgdWludDY0X3QqIGRzdCA9IHJlaW50ZXJwcmV0X2Nhc3Q8dWludDY0X3QqPihjICsgKHR4ICsgY3gpICogbiArIGN5ICsgdHkgKiA0KTsKICAgIGRzdFswXSA9IHBhY2tfdTE2eDhfdG9fdTY0KHNyY1swXSk7CiAgICBkc3RbMl0gPSBwYWNrX3UxNng4X3RvX3U2NChzcmNbMV0pOwogICAgZHN0WzRdID0gcGFja191MTZ4OF90b191NjQoc3JjWzJdKTsKICAgIGRzdFs2XSA9IHBhY2tfdTE2eDhfdG9fdTY0KHNyY1szXSk7Cn0KCl9fZ2xvYmFsX18gdm9pZCBwZXJtdXRlX2tlcm5lbChjb25zdCBfX2hpcF9mcDhfZTRtM19mbnV6KiBzcmMsIGNvbnN0IF9faGlwX2ZwOF9lNG0zX2ZudXoqIHNyYzEsIHVpbnQ2NF90KiBkc3QsCiAgICBzaXplX3QgbSwgc2l6ZV90IG4sIHNpemVfdCBrLCBzaXplX3QgY29sLCBzaXplX3QgYXMwLCBzaXplX3QgYXMxLCBzaXplX3QgYnMwLCBzaXplX3QgYnMxLCBzaXplX3QgbmV4dCkgewogICAgbmV4dCA9IGJsb2NrSWR4LnggPj0gbmV4dCA/IG5leHQgOiAwOwogICAgbSA9IG5leHQgPyBuIDogbTsKICAgIGFzMCA9IG5leHQgPyBiczAgOiBhczA7CiAgICBhczEgPSBuZXh0ID8gYnMxIDogYXMxOwogICAgc3JjID0gbmV4dCA/IHNyYzEgOiBzcmM7CiAgICBzaXplX3QgY3ggPSAoYmxvY2tJZHgueCAtIG5leHQpICogMzIgKyAodGhyZWFkSWR4LnggJiAzMSk7CiAgICBzaXplX3QgY3kgPSAoYmxvY2tJZHgueSAqIDIgKyAodGhyZWFkSWR4LnggPj4gNSAmIDEpKSAqIDg7CiAgICB1aW50NjRfdCB2YWx1ZSA9IDA7CiAgICBpZiAoY3ggPCBtICYmIGN5IDwgaykgewogICAgICAgIHNyYyArPSBjeSAqIGFzMSArIGN4ICogYXMwOwogICAgICAgIF9faGlwX2ZwOF9lNG0zX2ZudXogYnVmZmVyWzhdID0geyBzcmNbMF0sIHNyY1thczFdLCBzcmNbMiAqIGFzMV0sIHNyY1szICogYXMxXSwKICAgICAgICAgICAgc3JjWzQgKiBhczFdLCBzcmNbNSAqIGFzMV0sIHNyY1s2ICogYXMxXSwgc3JjWzcgKiBhczFdfTsKICAgICAgICB2YWx1ZSA9ICpyZWludGVycHJldF9jYXN0PHVpbnQ2NF90Kj4oYnVmZmVyKTsKICAgIH0KICAgIGRzdFtibG9ja0lkeC54ICogY29sICsgYmxvY2tJZHgueSAqIDY0ICsgdGhyZWFkSWR4LnhdID0gdmFsdWU7Cn0KCnZvaWQgZnA4X21tKHRvcmNoOjpUZW5zb3IgYSwgdG9yY2g6OlRlbnNvciBiLCB0b3JjaDo6VGVuc29yIGFzLCB0b3JjaDo6VGVuc29yIGJzLCB0b3JjaDo6VGVuc29yIGMpIHsKICAgIHNpemVfdCBtID0gYS5zaXplKDApOwogICAgc2l6ZV90IG4gPSBiLnNpemUoMCk7CiAgICBzaXplX3QgayA9IGEuc2l6ZSgxKTsKICAgIHNpemVfdCBuZXh0ID0gKG0gKyAxMjcpIC8gMTI4ICogNDsKICAgIHNpemVfdCByb3cgPSAobiArIDEyNykgLyAxMjggKiA0ICsgbmV4dDsKICAgIHNpemVfdCBjb2wgPSAoayArIDEyNykgLyAxMjggKiA1MTI7CiAgICBhdXRvIG9wdGlvbnMgPSB0b3JjaDo6VGVuc29yT3B0aW9ucygpLmR0eXBlKHRvcmNoOjprVUludDY0KS5kZXZpY2UodG9yY2g6OmtDVURBKTsKICAgIHRvcmNoOjpUZW5zb3IgY2FjaGUgPSB0b3JjaDo6ZW1wdHkocm93ICogY29sLCBvcHRpb25zKTsKICAgIHBlcm11dGVfa2VybmVsPDw8ZGltMyhyb3csIGNvbCAvIDY0KSwgZGltMyg2NCksIDAsIDA+Pj4oKF9faGlwX2ZwOF9lNG0zX2ZudXoqKWEuZGF0YV9wdHIoKSwgIChfX2hpcF9mcDhfZTRtM19mbnV6KiliLmRhdGFfcHRyKCksCiAgICAgICAgY2FjaGUuZGF0YV9wdHI8dWludDY0X3Q+KCksIG0sIG4sIGssIGNvbCwgYS5zdHJpZGUoMCksIGEuc3RyaWRlKDEpLCBiLnN0cmlkZSgwKSwgYi5zdHJpZGUoMSksIG5leHQpOwogICAgY3VzdG9tX2tlcm5lbDw8PGRpbTMoKG0rMzEpLzMyLCAobiszMSkvMzIpLCBkaW0zKDY0KSwgMCwgMD4+PiAoY2FjaGUuZGF0YV9wdHI8dWludDY0X3Q+KCksCiAgICAgICAgYXMuZGF0YV9wdHI8ZmxvYXQ+KCksIGJzLmRhdGFfcHRyPGZsb2F0PigpLCAoX19oaXBfYmZsb2F0MTYqKWMuZGF0YV9wdHIoKSwgbSwgbiwgaywgY29sLCBuZXh0ICogY29sKTsKfQoiIiIKCmltcG9ydCBvcwpvcy5lbnZpcm9uWyJDWFgiXSA9ICJjbGFuZysrIgpvcy5lbnZpcm9uWydQWVRPUkNIX1JPQ01fQVJDSCddID0gJ2dmeDk0MicKCm1vZHVsZSA9IGxvYWRfaW5saW5lKAogICAgbmFtZT0nZnA4X21tJywKICAgIGNwcF9zb3VyY2VzPVtDUFBfV1JBUFBFUl0sCiAgICBjdWRhX3NvdXJjZXM9W0NVREFfU1JDXSwKICAgIGZ1bmN0aW9ucz1bJ2ZwOF9tbSddLAogICAgdmVyYm9zZT1UcnVlLAogICAgZXh0cmFfY3VkYV9jZmxhZ3M9WyItLW9mZmxvYWQtYXJjaD1nZng5NDIiLCAiLXN0ZD1jKysyMCIsICItTzMiXSwKKQoKCmRlZiBjdXN0b21fa2VybmVsKGRhdGE6IGlucHV0X3QpIC0+IG91dHB1dF90OgogICAgYSwgYiwgYV9zY2FsZSwgYl9zY2FsZSwgYyA9IGRhdGEKICAgIG1vZHVsZS5mcDhfbW0oYSwgYiwgYV9zY2FsZSwgYl9zY2FsZSwgYykKICAgIHJldHVybiBj

Source code from GPU Mode and the KernelBot dataset · June 9 Researcher Reciprocity License v1.0

Best evidence level for this revision: reported

JSON