source stringclasses 13
values | gpu stringclasses 4
values | model stringclasses 13
values | kernel_family stringclasses 7
values | kernel_name stringclasses 328
values | dtype stringclasses 2
values | held_out bool 2
classes | M float64 1 16.4k ⌀ | N float64 8 248k ⌀ | K float64 2.05k 29.6k ⌀ | bs float64 1 1 ⌀ | seq float64 1 16.4k ⌀ | n_heads float64 8 64 ⌀ | head_dim float64 64 256 ⌀ | kv_heads float64 1 64 ⌀ | numel float64 64 531M ⌀ | op_type stringclasses 177
values | gpu_time_duration_ms float64 0 703 | launch_block_size float64 16 512 | launch_grid_size float64 1 1.04M | dram_bytes_sum float64 1.27 94.3B ⌀ | launch_registers_per_thread float64 16 255 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,777,216 | residual | 0.05011 | 128 | 32,768 | 51,420,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | other | 0.13478 | 512 | 528 | 52,000,000 | 25 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<c10::BFloat16, at::native::MaxOps<c10::BFloat16>, unsigned int, c10::BFloat16, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.03882 | 512 | 2,048 | 34,800,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,777,216 | residual | 0.04944 | 128 | 32,832 | 50,390,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void <unnamed>::softmax_warp_forward<c10::BFloat16, c10::BFloat16, float, (int)10, (bool)0, (bool)0>(T2 *, const T1 *, int, int, int, const bool *, int, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.07427 | 128 | 8,192 | 49,610,000 | 48 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 12)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const ... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,097,152 | cast | 0.0105 | 128 | 4,096 | 4,220,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | residual | 0.00678 | 128 | 1,440 | 5,900,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,474,560 | cast | 0.00896 | 128 | 2,880 | 2,980,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | rmsnorm | 0.00704 | 128 | 1,440 | 5,900,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | rmsnorm | 0.0152 | 512 | 32 | 5,960,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00336 | 128 | 1 | 7,170 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00362 | 128 | 1 | 8,960 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | mul | 0.00883 | 128 | 5,760 | 5,920,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,474,560 | mul | 0.01194 | 128 | 2,880 | 5,970,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | cast | 0.00624 | 128 | 1,440 | 5,900,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void <unnamed>::softmax_warp_forward<c10::BFloat16, c10::BFloat16, float, (int)2, (bool)0, (bool)0>(T2 *, const T1 *, int, int, int, const bool *, int, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,096 | other | 0.00371 | 128 | 8 | 11,010 | 21 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(long) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[la... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,096 | cast | 0.00438 | 128 | 8 | 17,150 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::compare_scalar_kernel<long>(at::TensorIteratorBase &, at::native::<unnamed>::OpType, T1)::[lambda(long) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,024 | other | 0.00323 | 128 | 2 | 26,110 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)2, at::native::<unnamed>::launch_clamp_scalar(at::TensorIteratorBase &, c10::Scalar, c10::Scalar, at::native::detail::ClampLimits)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(long) (instance 1)], std::array<char... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,024 | other | 0.00384 | 128 | 2 | 26,370 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_put_kernel_impl<at::native::OpaqueType<(int)8>>(at::TensorIterator &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00816 | 128 | 4 | 85,250 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_kernel_impl<at::native::OpaqueType<(int)8>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00787 | 128 | 4 | 81,920 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_kernel_impl<at::native::OpaqueType<(int)2>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00768 | 128 | 4 | 68,610 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_kernel_impl<at::native::OpaqueType<(int)8>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00755 | 128 | 4 | 81,410 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::vectorized_gather_kernel<(int)16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.00797 | 256 | 4,096 | 2,970,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 3)]::operator ()() const::[lambda(int) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int>,... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | cast | 0.00403 | 128 | 4 | 49,660 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::FillFunctor<int>, std::array<char *, (unsigned long)1>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | fill | 0.00291 | 128 | 1 | 4,100 | 16 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::vectorized_gather_kernel<(int)16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 6,291,456 | other | 0.01072 | 256 | 6,144 | 697,860 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | residual | 0.02138 | 128 | 11,520 | 55,520,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::<unnamed>::launch_clamp_scalar(at::TensorIteratorBase &, c10::Scalar, c10::Scalar, at::native::detail::ClampLimits)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lamb... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | other | 0.02074 | 128 | 11,520 | 24,330,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::<unnamed>::launch_clamp_scalar(at::TensorIteratorBase &, c10::Scalar, c10::Scalar, at::native::detail::ClampLimits)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lamb... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | other | 0.02077 | 128 | 11,520 | 24,230,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::AUnaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | mul | 0.00899 | 128 | 5,760 | 11,800,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::sigmoid_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | other | 0.01024 | 128 | 5,760 | 11,810,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | mul | 0.01194 | 128 | 5,760 | 23,880,000 | 40 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctorOnSelf_add<c10::BFloat16>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | residual | 0.00902 | 128 | 5,760 | 11,800,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | mul | 0.01219 | 128 | 5,760 | 23,740,000 | 40 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::vectorized_gather_kernel<(int)16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.008 | 256 | 4,096 | 245,500 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | residual | 0.0119 | 128 | 5,760 | 23,840,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | mul | 0.01974 | 128 | 11,520 | 11,830,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_kernel_impl<at::native::OpaqueType<(int)1>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00752 | 128 | 4 | 66,560 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::<unnamed>::masked_fill_kernel(at::TensorIterator &, const c10::Scalar &)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 12)]::operator ()() const::[lambda(c10::BFloat16, bool) (instance 1)]>(at:... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | other | 0.02118 | 128 | 11,520 | 11,820,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::vectorized_gather_kernel<(int)16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.00976 | 256 | 4,096 | 11,820,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void at::native::reduce_kernel<(int)128, (int)4, at::native::ReduceOp<c10::BFloat16, at::native::func_wrapper_t<c10::BFloat16, at::native::sum_functor<c10::BFloat16, float, c10::BFloat16>::operator ()(at::TensorIterator &)::[lambda(float, float) (instance 1)]>, unsigned int, c10::BFloat16, (int)4, (int)8>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,474,560 | other | 0.01478 | 128 | 2,880 | 11,870,000 | 54 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | residual | 0.00646 | 128 | 1,440 | 5,910,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,474,560 | cast | 0.00909 | 128 | 2,880 | 2,980,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | rmsnorm | 0.00704 | 128 | 1,440 | 5,900,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | rmsnorm | 0.01523 | 512 | 32 | 5,960,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00346 | 128 | 1 | 7,680 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00342 | 128 | 1 | 8,700 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | mul | 0.00858 | 128 | 5,760 | 5,920,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,474,560 | mul | 0.01197 | 128 | 2,880 | 5,970,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | cast | 0.00643 | 128 | 1,440 | 5,900,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | mul | 0.01066 | 128 | 2,048 | 2,160,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | mul | 0.01069 | 128 | 2,048 | 2,160,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | residual | 0.00522 | 128 | 1,024 | 4,200,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | mul | 0.01085 | 128 | 2,048 | 2,160,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | mul | 0.01075 | 128 | 2,048 | 2,160,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | residual | 0.00506 | 128 | 1,024 | 4,200,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | other | 0.01606 | 512 | 528 | 4,200,000 | 25 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | mul | 0.00669 | 128 | 256 | 321,790 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | mul | 0.00634 | 128 | 256 | 321,540 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | residual | 0.00355 | 128 | 128 | 531,200 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | mul | 0.00675 | 128 | 256 | 322,050 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | mul | 0.00662 | 128 | 256 | 322,300 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | residual | 0.00358 | 128 | 128 | 530,690 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | other | 0.00746 | 512 | 528 | 531,710 | 25 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | other | 0.0073 | 512 | 528 | 532,480 | 25 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | other | 0.00717 | 512 | 528 | 532,220 | 25 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 12)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const ... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,097,152 | cast | 0.00915 | 128 | 4,096 | 547,580 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 12)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const ... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,097,152 | cast | 0.00928 | 128 | 4,096 | 547,070 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::AUnaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 8,388,608 | mul | 0.02051 | 128 | 16,384 | 44,930,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,777,216 | residual | 0.05002 | 128 | 32,768 | 51,470,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | other | 0.13206 | 512 | 528 | 52,020,000 | 25 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<c10::BFloat16, at::native::MaxOps<c10::BFloat16>, unsigned int, c10::BFloat16, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.03846 | 512 | 2,048 | 34,880,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,777,216 | residual | 0.04944 | 128 | 32,832 | 50,240,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void <unnamed>::softmax_warp_forward<c10::BFloat16, c10::BFloat16, float, (int)10, (bool)0, (bool)0>(T2 *, const T1 *, int, int, int, const bool *, int, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.06461 | 128 | 8,192 | 49,580,000 | 48 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 12)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const ... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,097,152 | cast | 0.01034 | 128 | 4,096 | 4,220,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | residual | 0.0064 | 128 | 1,440 | 5,910,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,474,560 | cast | 0.00899 | 128 | 2,880 | 2,980,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | rmsnorm | 0.00666 | 128 | 1,440 | 5,900,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 92,160 | rmsnorm | 0.01514 | 512 | 32 | 5,960,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00342 | 128 | 1 | 7,420 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00346 | 128 | 1 | 8,700 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,949,120 | mul | 0.00832 | 128 | 5,760 | 5,910,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,474,560 | mul | 0.01219 | 128 | 2,880 | 5,970,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 737,280 | cast | 0.00662 | 128 | 1,440 | 5,900,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | reduce | void <unnamed>::softmax_warp_forward<c10::BFloat16, c10::BFloat16, float, (int)2, (bool)0, (bool)0>(T2 *, const T1 *, int, int, int, const bool *, int, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,096 | other | 0.00352 | 128 | 8 | 10,240 | 21 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(long) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[la... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,096 | cast | 0.00454 | 128 | 8 | 17,410 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::compare_scalar_kernel<long>(at::TensorIteratorBase &, at::native::<unnamed>::OpType, T1)::[lambda(long) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,024 | other | 0.00336 | 128 | 2 | 25,340 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)2, at::native::<unnamed>::launch_clamp_scalar(at::TensorIteratorBase &, c10::Scalar, c10::Scalar, at::native::detail::ClampLimits)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(long) (instance 1)], std::array<char... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,024 | other | 0.00355 | 128 | 2 | 26,370 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_put_kernel_impl<at::native::OpaqueType<(int)8>>(at::TensorIterator &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00854 | 128 | 4 | 81,150 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_kernel_impl<at::native::OpaqueType<(int)8>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00768 | 128 | 4 | 81,150 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_kernel_impl<at::native::OpaqueType<(int)2>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00739 | 128 | 4 | 68,610 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::index_elementwise_kernel<(int)128, (int)4, void at::native::gpu_index_kernel<void at::native::index_kernel_impl<at::native::OpaqueType<(int)8>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.00784 | 128 | 4 | 80,900 | 36 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::vectorized_gather_kernel<(int)16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 4,194,304 | other | 0.00803 | 256 | 4,096 | 2,970,000 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 3)]::operator ()() const::[lambda(int) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int>,... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | cast | 0.00387 | 128 | 4 | 49,150 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::FillFunctor<int>, std::array<char *, (unsigned long)1>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | fill | 0.00282 | 128 | 1 | 4,350 | 16 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | copy | void at::native::vectorized_gather_kernel<(int)16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 6,291,456 | other | 0.01082 | 256 | 6,144 | 795,390 | 32 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | residual | 0.0216 | 128 | 11,520 | 55,570,000 | 34 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::<unnamed>::launch_clamp_scalar(at::TensorIteratorBase &, c10::Scalar, c10::Scalar, at::native::detail::ClampLimits)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lamb... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | other | 0.0208 | 128 | 11,520 | 24,320,000 | 18 |
gpt-oss-20b_prefill | H100 | gpt-oss-20b | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::<unnamed>::launch_clamp_scalar(at::TensorIteratorBase &, c10::Scalar, c10::Scalar, at::native::detail::ClampLimits)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lamb... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 5,898,240 | other | 0.02064 | 128 | 11,520 | 24,150,000 | 18 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.