Compute- and bandwidth-bound CPU cases only. Per-call latency results are on the latency page.
8fec1ca70852cpu/cpu_ops 897 cases · 186 flaggedbaseline PyTorch · best first · 897 cases · case table, absolute times and raw report →
cpu/einsum 144 casesbaseline PyTorch · best first · 144 cases · case table, absolute times and raw report →
cpu/linalg_jvp_vjp 240 casesbaseline PyTorch · best first · 240 cases · case table, absolute times and raw report →
cpu/permutation 30 casesbaseline best reference · best first · 30 cases · case table, absolute times and raw report →
cyclic_15d_3, cyclic_15d_3, cyclic_15d_3, memcpy_24d_contiguous, memcpy_24d_contiguous, memcpy_24d_contiguous, reverse_15d_3, reverse_15d_3, reverse_15d_3, reverse_23d_2, reverse_23d_2, reverse_23d_2, rotation_6d_32_32_32_32_16_16, rotation_6d_32_32_32_32_16_16, rotation_6d_32_32_32_32_16_16, tn_light_415_24d_contiguous_same_perm, tn_light_415_24d_contiguous_same_perm, tn_light_415_24d_contiguous_same_perm, tn_light_415_24d_scattered_to_colmajor, tn_light_415_24d_scattered_to_colmajor, tn_light_415_24d_scattered_to_colmajor, transpose_2d_2048, transpose_2d_2048, transpose_2d_2048, transpose_3d_256_102, transpose_3d_256_102, transpose_3d_256_102, transpose_3d_256_201, transpose_3d_256_201, transpose_3d_256_201
cpu/public_api 1014 cases · 244 flaggedbaseline PyTorch · best first · 1014 cases · case table, absolute times and raw report →
add_into, add_into, add_into, broadcast_in_dim_view, broadcast_in_dim_view, broadcast_in_dim_view, conj_into, conj_into, conj_into, copy_read_into, copy_read_into, copy_read_into, div_into, div_into, div_into, dot_general_read_into, dot_general_read_into, dot_general_read_into, dot_general_read_into_accum, dot_general_read_into_accum, dot_general_read_into_accum, mul_into, mul_into, mul_into, neg_into, neg_into, neg_into, rank_revealing_qr, rank_revealing_qr, rank_revealing_qr, rank_revealing_qr, rank_revealing_qr, rank_revealing_qr, rank_revealing_qr_deficient, rank_revealing_qr_deficient, rank_revealing_qr_deficient, rank_revealing_qr_deficient, rank_revealing_qr_deficient, rank_revealing_qr_deficient, reduce_sum_squares_axis0, reduce_sum_squares_axis0, reduce_sum_squares_axis0, reshape_view, reshape_view, reshape_view, slice_view, slice_view, slice_view, solve_read_into, solve_read_into, solve_read_into, solve_read_into, solve_read_into, solve_read_into, sub_into, sub_into, sub_into, transpose_view, transpose_view, transpose_view
cpu/fft (collection date unrecorded)cpu/overhead_smoke (2026-10-28)cpu/session_matrix (collection date unrecorded)cpu/small_work (collection date unrecorded)