This repository is part of NUMCL.

Usage

./report.sh [test]

test specifies the test that needs to be rerun and displayed. It is one of:

• 0template
• 1access
• 1allocation
• 2linarg
• 3arith
• 4concat
• 5math

Dependencies

numpy: pip3 install numpy benchmarker or: conda env create -f environment.yml

numcl: ros install numcl/constantfold numcl/gtype numcl/specialized-function numcl/numcl trivial-benchmark

general: awk, make, other general unix tools

Author, License, Copyright

Masataro Asai ([email protected])

Licensed under LGPL v3.

Copyright (c) 2019 IBM Corporation

Logs

2020/6/9 thinkpad T480 SBCL 2.0.5

title	numcl@d296e3	numcl@0980fc	numcl@d296e3/numcl@0980fc	numpy@	numcl@0980fc/numpy@
0template/dummy/0	0.0040	0.0030	1.333	0.8939	0.003356
1access/read-range/0	0.0200	0.0180	1.111	0.0004	45
1access/read-range/1	0.0230	0.0140	1.643	0.0004	35
1access/read-range/2	0.2990	0.1320	2.265	0.0004	330
1access/read-range/3	0.2960	0.1770	1.672	0.0004	442.5
1access/read-range/4	0.2900	0.1910	1.518	0.0004	477.5
1access/read/0	0.0020	0.0020	1	0.0003	6.667
1access/read/1	0.0010	0.0020	0.5	0.0004	5
1access/read/2	0.0630	0.0430	1.465	0.0010	43
1access/read/3	0.0670	0.0470	1.426	0.0011	42.73
1access/read/4	0.0620	0.0490	1.265	0.0005	98
1access/write-batch/0	0.0170	0.0160	1.062	0.0009	17.78
1access/write-batch/1	0.0520	0.0510	1.02	0.0088	5.795
1access/write-range/0	0.0200	0.0220	0.9091	0.0008	27.5
1access/write-range/1	0.0200	0.0200	1	0.0008	25
1access/write-range/2	0.0380	0.0390	0.9744	0.0009	43.33
1access/write-range/3	0.0570	0.0470	1.213	0.0009	52.22
1access/write-range/4	0.0560	0.0480	1.167	0.0009	53.33
1access/write/0	0.0210	0.0210	1	0.0003	70
1access/write/1	0.0190	0.0190	1	0.0008	23.75
1access/write/2	0.0240	0.0250	0.96	0.0007	35.71
1access/write/3	0.0270	0.0280	0.9643	0.0007	40
1access/write/4	0.0380	0.0290	1.31	0.0007	41.43
1allocation/arange/float32	0.0000	0.0000	N/A	0.0001	0
1allocation/arange/float64	0.0010	0.0000	N/A	0.0002	0
1allocation/arange/int16	0.0000	0.0000	N/A	0.0001	0
1allocation/arange/int32	0.0000	0.0000	N/A	0.0001	0
1allocation/arange/int64	0.0000	0.0000	N/A	0.0001	0
1allocation/arange/int8	0.0000	0.0000	N/A	0.0001	0
1allocation/copy/bool	0.0040	0.0020	2	0.0572	0.03497
1allocation/copy/float32	0.1390	0.1240	1.121	0.0554	2.238
1allocation/copy/float64	0.3010	0.2840	1.06	0.0516	5.504
1allocation/copy/int16	0.0550	0.0550	1	0.0527	1.044
1allocation/copy/int32	0.1420	0.1250	1.136	0.0526	2.376
1allocation/copy/int64	0.3520	0.2880	1.222	0.0541	5.323
1allocation/copy/int8	0.0280	0.0280	1	0.0517	0.5416
1allocation/empty/bool	0.0020	0.0000	N/A	0.0001	0
1allocation/empty/float32	0.0520	0.0480	1.083	0.0001	480
1allocation/empty/float64	0.1040	0.0890	1.169	0.0002	445
1allocation/empty/int16	0.0260	0.0220	1.182	0.0001	220
1allocation/empty/int32	0.0520	0.0450	1.156	0.0001	450
1allocation/empty/int64	0.1070	0.0880	1.216	0.0002	440
1allocation/empty/int8	0.0130	0.0120	1.083	0.0001	120
1allocation/ones/bool	0.0090	0.0090	1	0.0045	2
1allocation/ones/float32	0.1120	0.1020	1.098	0.0261	3.908
1allocation/ones/float64	0.2210	0.2050	1.078	0.0831	2.467
1allocation/ones/int16	0.1020	0.1070	0.9533	0.0120	8.917
1allocation/ones/int32	0.1930	0.2150	0.8977	0.0453	4.746
1allocation/ones/int64	0.2890	0.2560	1.129	0.0864	2.963
1allocation/ones/int8	0.0680	0.0710	0.9577	0.0033	21.52
1allocation/zeros/bool	0.0020	0.0030	0.6667	0.0038	0.7895
1allocation/zeros/float32	0.1000	0.1060	0.9434	0.0214	4.953
1allocation/zeros/float64	0.2140	0.1930	1.109	0.0684	2.822
1allocation/zeros/int16	0.0530	0.0490	1.082	0.0078	6.282
1allocation/zeros/int32	0.1050	0.0980	1.071	0.0245	4
1allocation/zeros/int64	0.2210	0.2000	1.105	0.0656	3.049
1allocation/zeros/int8	0.0200	0.0220	0.9091	0.0035	6.286
2linarg/diag/builtin	0.0060	0.0080	0.75	0.0008	10
2linarg/diag/einsum	0.0070	0.0050	1.4	0.0004	12.5
2linarg/eye/builtin	0.0000	0.0000	N/A	0.0013	0
2linarg/eye/naive	0.1760	0.1690	1.041	0.0039	43.33
2linarg/gemm-large/builtin	133.7070	97.6810	1.369	2.3147	42.2
2linarg/gemm-large/einsum	143.9220	94.3460	1.525	30.3643	3.107
2linarg/gemm/builtin	0.1480	0.0760	1.947	0.0120	6.333
2linarg/gemm/einsum	0.1410	0.0870	1.621	0.0411	2.117
2linarg/inner/builtin	0.0060	0.0060	1	0.0004	15
2linarg/inner/einsum	0.0050	0.0080	0.625	0.0009	8.889
2linarg/outer/builtin	0.0080	0.0090	0.8889	0.0041	2.195
2linarg/outer/einsum	0.0070	0.0100	0.7	0.0031	3.226
2linarg/tri/builtin	0.0170	0.0180	0.9444	0.0069	2.609
2linarg/tril/builtin	0.0230	0.0270	0.8519	0.0082	3.293
2linarg/triu/builtin	0.0250	0.0270	0.9259	0.0081	3.333
2linarg/vander/builtin	0.0620	0.0630	0.9841	0.0097	6.495
2linarg/vdot/builtin	0.0060	0.0070	0.8571	0.0004	17.5
2linarg/vdot/einsum	0.0060	0.0080	0.75	0.0013	6.154
3arith/add_cd_cd	N/A	N/A	N/A	0.0015	N/A
3arith/add_cs_cs	N/A	N/A	N/A	0.0017	N/A
3arith/add_d_d	0.0290	0.0180	1.611	0.0007	25.71
3arith/add_d_i32	0.0290	0.0200	1.45	0.0014	14.29
3arith/add_i16_i16	0.0260	0.0190	1.368	0.0003	63.33
3arith/add_i1_i1	0.0390	0.0280	1.393	0.0002	140
3arith/add_i32_i32	0.0300	0.0200	1.5	0.0003	66.67
3arith/add_i64_i64	0.0330	0.0190	1.737	0.0006	31.67
3arith/add_i8_i8	0.0320	0.0170	1.882	0.0002	85
3arith/add_s_d	0.0300	0.0190	1.579	0.0014	13.57
3arith/add_s_i32	0.0260	0.0170	1.529	0.0019	8.947
3arith/add_s_s	0.0270	0.0180	1.5	0.0003	60
3arith/fma_cd_cd_cd	N/A	N/A	N/A	0.0043	N/A
3arith/fma_cs_cs_cs	N/A	N/A	N/A	0.0040	N/A
3arith/fma_d_d_d	0.0640	0.0350	1.829	0.0012	29.17
3arith/fma_d_i32_d	0.0660	0.0360	1.833	0.0021	17.14
3arith/fma_i16_i16_i16	0.0760	0.0390	1.949	0.0004	97.5
3arith/fma_i1_i1_i1	0.0780	0.0500	1.56	0.0003	166.7
3arith/fma_i32_i32_i32	0.0900	0.0460	1.957	0.0006	76.67
3arith/fma_i64_i64_i64	0.0740	0.0380	1.947	0.0013	29.23
3arith/fma_i8_i8_i8	0.0720	0.0390	1.846	0.0003	130
3arith/fma_s_i32_d	N/A	N/A	N/A	0.0027	N/A
3arith/fma_s_i32_s	0.0600	0.0370	1.622	0.0034	10.88
3arith/fma_s_s_s	0.0620	0.0330	1.879	0.0006	55
3arith/mul_cd_cd	N/A	N/A	N/A	0.0022	N/A
3arith/mul_cs_cs	N/A	N/A	N/A	0.0022	N/A
3arith/mul_d_d	0.0560	0.0200	2.8	0.0005	40
3arith/mul_d_i32	0.0310	0.0160	1.938	0.0014	11.43
3arith/mul_i16_i16	0.0280	0.0220	1.273	0.0002	110
3arith/mul_i1_i1	0.0390	0.0210	1.857	0.0001	210
3arith/mul_i32_i32	0.0330	0.0160	2.062	0.0003	53.33
3arith/mul_i64_i64	0.0330	0.0160	2.062	0.0007	22.86
3arith/mul_i8_i8	0.0250	0.0190	1.316	0.0002	95
3arith/mul_s_d	0.0560	0.0130	4.308	0.0014	9.286
3arith/mul_s_i32	0.0290	0.0150	1.933	0.0020	7.5
3arith/mul_s_s	0.0320	0.0140	2.286	0.0003	46.67
4concat/concatenate/0	0.0040	0.0050	0.8	0.0003	16.67
4concat/concatenate/1	0.0050	0.0090	0.5556	0.0003	30
4concat/concatenate/2	0.0270	0.0360	0.75	0.0004	90
4concat/stack/0	0.0030	0.0050	0.6	0.0009	5.556
4concat/stack/1	0.0050	0.0070	0.7143	0.0009	7.778
4concat/stack/2	0.0300	0.0400	0.75	0.0010	40
5math/acos/0	1.5480	1.5670	0.9879	0.5140	3.049
5math/asin/0	1.8250	1.8550	0.9838	0.4661	3.98
5math/atan/0	1.1710	1.2180	0.9614	0.7377	1.651
5math/cos/0	0.8620	1.1550	0.7463	0.4401	2.624
5math/cosh/0	0.7380	0.7710	0.9572	0.5496	1.403
5math/exp/0	1.2320	0.9950	1.238	0.8445	1.178
5math/log/0	1.1480	2.4500	0.4686	0.7065	3.468
5math/sin/0	1.1890	1.1290	1.053	0.6603	1.71
5math/sinh/0	1.0450	0.9690	1.078	0.7570	1.28
5math/tan/0	1.0480	1.2420	0.8438	0.6351	1.956
5math/tanh/0	0.6400	0.5330	1.201	0.4663	1.143

2020/2/26 thinkpad T480

title	numcl@0980fc	numcl@202820	numcl@0980fc/numcl@202820	numpy	numcl@202820/numpy
	(old master)	(new master)	(old/new)
0template/dummy/0	0.0040	0.0030	1.333	1.3115	0.002287
1access/read/0	0.0020	0.0020	1	0.0003	6.667
1access/read/1	0.0020	0.0020	1	0.0003	6.667
1access/read/2	0.0520	0.0520	1	0.0004	130
1access/read/3	0.0460	0.0530	0.8679	0.0004	132.5
1access/read/4	0.0460	0.0500	0.92	0.0004	125
1access/read-range/0	0.0170	0.0180	0.9444	0.0004	45
1access/read-range/1	0.0180	0.0180	1	0.0004	45
1access/read-range/2	0.1390	0.1630	0.8528	0.0004	407.5
1access/read-range/3	0.1750	0.1700	1.029	0.0005	340
1access/read-range/4	0.1620	0.1570	1.032	0.0004	392.5
1access/write/0	0.0210	0.0210	1	0.0002	105
1access/write/1	0.0200	0.0170	1.176	0.0007	24.29
1access/write/2	0.0250	0.0230	1.087	0.0007	32.86
1access/write/3	0.0270	0.0290	0.931	0.0007	41.43
1access/write/4	0.0310	0.0300	1.033	0.0008	37.5
1access/write-batch/0	0.0130	0.0100	1.3	0.0008	12.5
1access/write-batch/1	0.0460	0.0410	1.122	0.0080	5.125
1access/write-range/0	0.0250	0.0200	1.25	0.0007	28.57
1access/write-range/1	0.0260	0.0200	1.3	0.0008	25
1access/write-range/2	0.0490	0.0400	1.225	0.0009	44.44
1access/write-range/3	0.0490	0.0450	1.089	0.0009	50
1access/write-range/4	0.0520	0.0540	0.963	0.0009	60
1allocation/arange/float32	0.0020	0.0040	0.5	0.0002	20
1allocation/arange/float64	0.0040	0.0030	1.333	0.0002	15
1allocation/arange/int16	0.0010	0.0000	N/A	0.0001	0
1allocation/arange/int32	0.0030	0.0020	1.5	0.0001	20
1allocation/arange/int64	0.0050	0.0050	1	0.0001	50
1allocation/arange/int8	0.0000	0.0000	N/A	0.0001	0
1allocation/copy/bool	0.0030	0.0030	1	0.0706	0.04249
1allocation/copy/float32	0.0940	0.1000	0.94	0.0575	1.739
1allocation/copy/float64	0.2660	0.2710	0.9815	0.0606	4.472
1allocation/copy/int16	0.0380	0.0450	0.8444	0.0800	0.5625
1allocation/copy/int32	0.1010	0.1020	0.9902	0.0711	1.435
1allocation/copy/int64	0.2670	0.2720	0.9816	0.0607	4.481
1allocation/copy/int8	0.0180	0.0180	1	0.0746	0.2413
1allocation/empty/bool	0.0010	0.0010	1	0.0001	10
1allocation/empty/float32	0.0390	0.0370	1.054	0.0001	370
1allocation/empty/float64	0.0790	0.0780	1.013	0.0002	390
1allocation/empty/int16	0.0190	0.0190	1	0.0001	190
1allocation/empty/int32	0.0370	0.0370	1	0.0001	370
1allocation/empty/int64	0.0780	0.0770	1.013	0.0002	385
1allocation/empty/int8	0.0100	0.0090	1.111	0.0001	90
1allocation/ones/bool	0.0030	0.0020	1.5	0.0035	0.5714
1allocation/ones/float32	0.0720	0.0670	1.075	0.1356	0.4941
1allocation/ones/float64	0.1920	0.1960	0.9796	0.1214	1.614
1allocation/ones/int16	0.0580	0.0600	0.9667	0.0084	7.143
1allocation/ones/int32	0.1240	0.1230	1.008	0.0277	4.44
1allocation/ones/int64	0.2440	0.2320	1.052	0.1238	1.874
1allocation/ones/int8	0.0440	0.0410	1.073	0.0027	15.19
1allocation/zeros/bool	0.0020	0.0020	1	0.0043	0.4651
1allocation/zeros/float32	0.0690	0.1000	0.69	0.0294	3.401
1allocation/zeros/float64	0.1970	0.1920	1.026	0.0941	2.04
1allocation/zeros/int16	0.0300	0.0300	1	0.0106	2.83
1allocation/zeros/int32	0.0730	0.0720	1.014	0.0315	2.286
1allocation/zeros/int64	0.2090	0.2150	0.9721	0.0810	2.654
1allocation/zeros/int8	0.0160	0.0190	0.8421	0.0038	5
2linarg/diag/builtin	0.0050	0.0050	1	0.0004	12.5
2linarg/diag/einsum	0.0050	0.0050	1	0.0002	25
2linarg/eye/builtin	0.0000	0.0000	N/A	0.0007	0
2linarg/eye/naive	0.1790	0.1900	0.9421	0.0019	100
2linarg/gemm/builtin	0.0720	0.1310	0.5496	0.0124	10.56
2linarg/gemm/einsum	0.0750	0.1340	0.5597	0.0457	2.932
2linarg/gemm-large/builtin	63.9760	124.4860	0.5139	2.0091	61.96
2linarg/gemm-large/einsum	64.4450	122.9670	0.5241	28.4868	4.317
2linarg/inner/builtin	0.0040	0.0040	1	0.0004	10
2linarg/inner/einsum	0.0040	0.0040	1	0.0012	3.333
2linarg/outer/builtin	0.0070	0.0070	1	0.0045	1.556
2linarg/outer/einsum	0.0090	0.0050	1.8	0.0030	1.667
2linarg/tri/builtin	0.0130	0.0130	1	0.0077	1.688
2linarg/tril/builtin	0.0210	0.0210	1	0.0083	2.53
2linarg/triu/builtin	0.0220	0.0210	1.048	0.0081	2.593
2linarg/vander/builtin	0.0510	0.0490	1.041	0.0091	5.385
2linarg/vdot/builtin	0.0070	0.0050	1.4	0.0004	12.5
2linarg/vdot/einsum	0.0070	0.0040	1.75	0.0013	3.077
3arith/add_cd_cd	N/A	N/A	N/A	0.0016	N/A
3arith/add_cs_cs	N/A	N/A	N/A	0.0012	N/A
3arith/add_d_d	0.0120	0.0110	1.091	0.0004	27.5
3arith/add_d_i32	0.0130	0.0150	0.8667	0.0011	13.64
3arith/add_i16_i16	0.0120	0.0130	0.9231	0.0002	65
3arith/add_i1_i1	0.0170	0.0190	0.8947	0.0001	190
3arith/add_i32_i32	0.0150	0.0140	1.071	0.0003	46.67
3arith/add_i64_i64	0.0110	0.0170	0.6471	0.0006	28.33
3arith/add_i8_i8	0.0110	0.0140	0.7857	0.0001	140
3arith/add_s_d	0.0130	0.0120	1.083	0.0012	10
3arith/add_s_i32	0.0130	0.0110	1.182	0.0017	6.471
3arith/add_s_s	0.0110	0.0130	0.8462	0.0002	65
3arith/fma_cd_cd_cd	N/A	N/A	N/A	0.0085	N/A
3arith/fma_cs_cs_cs	N/A	N/A	N/A	0.0024	N/A
3arith/fma_d_d_d	0.0230	0.0230	1	0.0009	25.56
3arith/fma_d_i32_d	0.0230	0.0250	0.92	0.0014	17.86
3arith/fma_i16_i16_i16	0.0280	0.0280	1	0.0003	93.33
3arith/fma_i1_i1_i1	0.0380	0.0400	0.95	0.0002	200
3arith/fma_i32_i32_i32	0.0270	0.0290	0.931	0.0005	58
3arith/fma_i64_i64_i64	0.0290	0.0310	0.9355	0.0010	31
3arith/fma_i8_i8_i8	0.0230	0.0270	0.8519	0.0002	135
3arith/fma_s_i32_d	N/A	N/A	N/A	0.0019	N/A
3arith/fma_s_i32_s	0.0220	0.0240	0.9167	0.0024	10
3arith/fma_s_s_s	0.0290	0.0240	1.208	0.0004	60
3arith/mul_cd_cd	N/A	N/A	N/A	0.0018	N/A
3arith/mul_cs_cs	N/A	N/A	N/A	0.0017	N/A
3arith/mul_d_d	0.0130	0.0120	1.083	0.0005	24
3arith/mul_d_i32	0.0120	0.0130	0.9231	0.0010	13
3arith/mul_i16_i16	0.0120	0.0120	1	0.0002	60
3arith/mul_i1_i1	0.0170	0.0190	0.8947	0.0001	190
3arith/mul_i32_i32	0.0110	0.0150	0.7333	0.0003	50
3arith/mul_i64_i64	0.0110	0.0130	0.8462	0.0007	18.57
3arith/mul_i8_i8	0.0130	0.0110	1.182	0.0002	55
3arith/mul_s_d	0.0120	0.0120	1	0.0011	10.91
3arith/mul_s_i32	0.0130	0.0120	1.083	0.0015	8
3arith/mul_s_s	0.0120	0.0130	0.9231	0.0003	43.33
4concat/concatenate/0	0.0020	0.0020	1	0.0004	5
4concat/concatenate/1	0.0050	0.0060	0.8333	0.0003	20
4concat/concatenate/2	0.0290	0.0250	1.16	0.0004	62.5
4concat/stack/0	0.0040	0.0040	1	0.0010	4
4concat/stack/1	0.0060	0.0060	1	0.0010	6
4concat/stack/2	0.0300	0.0320	0.9375	0.0011	29.09
5math/acos/0	1.0610	1.1960	0.8871	0.4452	2.686
5math/asin/0	1.0240	1.1660	0.8782	0.4621	2.523
5math/atan/0	0.7790	0.8570	0.909	0.6743	1.271
5math/cos/0	0.7390	0.6740	1.096	0.4278	1.576
5math/cosh/0	0.5730	0.7410	0.7733	0.5246	1.413
5math/exp/0	0.7900	0.8990	0.8788	0.7668	1.172
5math/log/0	1.5530	1.8140	0.8561	0.6683	2.714
5math/sin/0	0.7880	0.8120	0.9704	0.8215	0.9884
5math/sinh/0	0.7760	1.1640	0.6667	0.7041	1.653
5math/tan/0	0.9550	0.8880	1.075	0.5731	1.549
5math/tanh/0	0.4500	0.6040	0.745	0.5101	1.184

2019/12/08 RYZEN1700

title	numcl @ 9602407	numpy 1.14.2	cl/py
0template/dummy/0	0.0030	0.8337	0.003598
1access/read-range/0	0.0190	0.0004	47.5
1access/read-range/1	0.0190	0.0004	47.5
1access/read-range/2	0.1840	0.0004	460
1access/read-range/3	0.2250	0.0004	562.5
1access/read-range/4	0.2230	0.0005	446
1access/read/0	0.0020	0.0002	10
1access/read/1	0.0030	0.0003	10
1access/read/2	0.0610	0.0003	203.3
1access/read/3	0.0520	0.0004	130
1access/read/4	0.0550	0.0003	183.3
1access/write-batch/0	0.0140	0.0010	14
1access/write-batch/1	0.0580	0.0089	6.517
1access/write-range/0	0.0220	0.0008	27.5
1access/write-range/1	0.0200	0.0012	16.67
1access/write-range/2	0.0390	0.0011	35.45
1access/write-range/3	0.0420	0.0012	35
1access/write-range/4	0.0460	0.0012	38.33
1access/write/0	0.0180	0.0002	90
1access/write/1	0.0180	0.0008	22.5
1access/write/2	0.0250	0.0008	31.25
1access/write/3	0.0280	0.0007	40
1access/write/4	0.0310	0.0007	44.29
1allocation/arange/float32	0.0040	0.0001	40
1allocation/arange/float64	0.0040	0.0001	40
1allocation/arange/int16	0.0020	0.0001	20
1allocation/arange/int32	0.0020	0.0001	20
1allocation/arange/int64	0.0040	0.0001	40
1allocation/arange/int8	0.0000	0.0001	0
1allocation/copy/bool	0.0040	0.0628	0.06369
1allocation/copy/float32	0.1260	0.0583	2.161
1allocation/copy/float64	0.3200	0.0646	4.954
1allocation/copy/int16	0.0510	0.0572	0.8916
1allocation/copy/int32	0.1260	0.0636	1.981
1allocation/copy/int64	0.3450	0.0551	6.261
1allocation/copy/int8	0.0260	0.0582	0.4467
1allocation/empty/bool	0.0030	0.0001	30
1allocation/empty/float32	0.0700	0.0001	700
1allocation/empty/float64	0.1420	0.0001	1420
1allocation/empty/int16	0.0400	0.0001	400
1allocation/empty/int32	0.0680	0.0001	680
1allocation/empty/int64	0.1430	0.0001	1430
1allocation/empty/int8	0.0190	0.0001	190
1allocation/ones/bool	0.0030	0.0030	1
1allocation/ones/float32	0.1210	0.0276	4.384
1allocation/ones/float64	0.2720	0.1137	2.392
1allocation/ones/int16	0.0460	0.0060	7.667
1allocation/ones/int32	0.1450	0.0390	3.718
1allocation/ones/int64	0.2800	0.1328	2.108
1allocation/ones/int8	0.0230	0.0024	9.583
1allocation/zeros/bool	0.0030	0.0025	1.2
1allocation/zeros/float32	0.1170	0.0280	4.179
1allocation/zeros/float64	0.2790	0.1277	2.185
1allocation/zeros/int16	0.0520	0.0049	10.61
1allocation/zeros/int32	0.1220	0.0238	5.126
1allocation/zeros/int64	0.2980	0.1259	2.367
1allocation/zeros/int8	0.0260	0.0022	11.82
2linarg/diag/builtin	0.0050	0.0008	6.25
2linarg/diag/einsum	0.0050	0.0003	16.67
2linarg/eye/builtin	0.0000	0.0010	0
2linarg/eye/naive	0.1740	0.0035	49.71
2linarg/gemm-large/builtin	44.3280	2.3508	18.86
2linarg/gemm-large/einsum	46.3890	23.8243	1.947
2linarg/gemm/builtin	0.0490	0.0259	1.892
2linarg/gemm/einsum	0.0480	0.0620	0.7742
2linarg/inner/builtin	0.0050	0.0001	50
2linarg/inner/einsum	0.0050	0.0003	16.67
2linarg/outer/builtin	0.0060	0.0022	2.727
2linarg/outer/einsum	0.0080	0.0013	6.154
2linarg/tri/builtin	0.0140	0.0036	3.889
2linarg/tril/builtin	0.0200	0.0036	5.556
2linarg/triu/builtin	0.0190	0.0035	5.429
2linarg/vander/builtin	0.0540	0.0068	7.941
2linarg/vdot/builtin	0.0070	0.0001	70
2linarg/vdot/einsum	0.0070	0.0004	17.5
3arith/add_cd_cd	N/A	0.0013	N/A
3arith/add_cs_cs	N/A	0.0010	N/A
3arith/add_d_d	0.0140	0.0003	46.67
3arith/add_d_i32	0.0140	0.0011	12.73
3arith/add_i16_i16	0.0130	0.0002	65
3arith/add_i1_i1	0.0140	0.0001	140
3arith/add_i32_i32	0.0160	0.0003	53.33
3arith/add_i64_i64	0.0150	0.0005	30
3arith/add_i8_i8	0.0120	0.0002	60
3arith/add_s_d	0.0140	0.0009	15.56
3arith/add_s_i32	0.0130	0.0016	8.125
3arith/add_s_s	0.0120	0.0002	60
3arith/fma_cd_cd_cd	N/A	0.0028	N/A
3arith/fma_cs_cs_cs	N/A	0.0024	N/A
3arith/fma_d_d_d	0.0250	0.0008	31.25
3arith/fma_d_i32_d	0.0250	0.0015	16.67
3arith/fma_i16_i16_i16	0.0260	0.0004	65
3arith/fma_i1_i1_i1	0.0270	0.0003	90
3arith/fma_i32_i32_i32	0.0260	0.0007	37.14
3arith/fma_i64_i64_i64	0.0260	0.0014	18.57
3arith/fma_i8_i8_i8	0.0250	0.0003	83.33
3arith/fma_s_i32_d	N/A	0.0018	N/A
3arith/fma_s_i32_s	0.0240	0.0022	10.91
3arith/fma_s_s_s	0.0230	0.0005	46
3arith/mul_cd_cd	N/A	0.0018	N/A
3arith/mul_cs_cs	N/A	0.0016	N/A
3arith/mul_d_d	0.0140	0.0005	28
3arith/mul_d_i32	0.0130	0.0009	14.44
3arith/mul_i16_i16	0.0130	0.0002	65
3arith/mul_i1_i1	0.0130	0.0002	65
3arith/mul_i32_i32	0.0150	0.0004	37.5
3arith/mul_i64_i64	0.0140	0.0009	15.56
3arith/mul_i8_i8	0.0120	0.0002	60
3arith/mul_s_d	0.0140	0.0009	15.56
3arith/mul_s_i32	0.0130	0.0013	10
3arith/mul_s_s	0.0120	0.0003	40
4concat/concatenate/0	0.0030	0.0003	10
4concat/concatenate/1	0.0050	0.0003	16.67
4concat/concatenate/2	0.0290	0.0004	72.5
4concat/stack/0	0.0030	0.0008	3.75
4concat/stack/1	0.0060	0.0007	8.571
4concat/stack/2	0.0350	0.0009	38.89
5math/acos/0	1.0010	0.5164	1.938
5math/asin/0	0.9530	0.4243	2.246
5math/atan/0	0.9750	0.8358	1.167
5math/cos/0	1.2620	0.7917	1.594
5math/cosh/0	0.4600	0.4140	1.111
5math/exp/0	0.9710	0.8494	1.143
5math/log/0	1.4590	0.4535	3.217
5math/sin/0	1.1580	0.8832	1.311
5math/sinh/0	0.5890	0.4834	1.218
5math/tan/0	1.2220	0.7911	1.545
5math/tanh/0	0.3860	0.3471	1.112