devtools/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2026-04-10 11:34:33 +08:00
Code Issues Packages Projects Releases Wiki Activity
10,664 Commits 15 Branches 93 Tags
f59bed7a133322955dac03f3654def706aff3ba6
Commit Graph

72 Commits

Author SHA1 Message Date
Eugene Zhulenev
e9f0eb8a5e Add masked_store_available to unpacket_traits 2019-05-02 14:52:58 -07:00
Anuj Rawat
8c7a6feb8e Adding lowlevel APIs for optimized RHS packet load in TensorFlow 
SpatialConvolution

Low-level APIs are added in order to optimized packet load in gemm_pack_rhs
in TensorFlow SpatialConvolution. The optimization is for scenario when a
packet is split across 2 adjacent columns. In this case we read it as two
'partial' packets and then merge these into 1. Currently this only works for
Packet16f (AVX512) and Packet8f (AVX2). We plan to add this for other
packet types (such as Packet8d) also.

This optimization shows significant speedup in SpatialConvolution with
certain parameters. Some examples are below.

Benchmark parameters are specified as:
Batch size, Input dim, Depth, Num of filters, Filter dim

Speedup numbers are specified for number of threads 1, 2, 4, 8, 16.

AVX512:

Parameters                  | Speedup (Num of threads: 1, 2, 4, 8, 16)
----------------------------|------------------------------------------
128,   24x24,  3, 64,   5x5 |2.18X, 2.13X, 1.73X, 1.64X, 1.66X
128,   24x24,  1, 64,   8x8 |2.00X, 1.98X, 1.93X, 1.91X, 1.91X
 32,   24x24,  3, 64,   5x5 |2.26X, 2.14X, 2.17X, 2.22X, 2.33X
128,   24x24,  3, 64,   3x3 |1.51X, 1.45X, 1.45X, 1.67X, 1.57X
 32,   14x14, 24, 64,   5x5 |1.21X, 1.19X, 1.16X, 1.70X, 1.17X
128, 128x128,  3, 96, 11x11 |2.17X, 2.18X, 2.19X, 2.20X, 2.18X

AVX2:

Parameters                  | Speedup (Num of threads: 1, 2, 4, 8, 16)
----------------------------|------------------------------------------
128,   24x24,  3, 64,   5x5 | 1.66X, 1.65X, 1.61X, 1.56X, 1.49X
 32,   24x24,  3, 64,   5x5 | 1.71X, 1.63X, 1.77X, 1.58X, 1.68X
128,   24x24,  1, 64,   5x5 | 1.44X, 1.40X, 1.38X, 1.37X, 1.33X
128,   24x24,  3, 64,   3x3 | 1.68X, 1.63X, 1.58X, 1.56X, 1.62X
128, 128x128,  3, 96, 11x11 | 1.36X, 1.36X, 1.37X, 1.37X, 1.37X

In the higher level benchmark cifar10, we observe a runtime improvement
of around 6% for AVX512 on Intel Skylake server (8 cores).

On lower level PackRhs micro-benchmarks specified in TensorFlow
tensorflow/core/kernels/eigen_spatial_convolutions_test.cc, we observe
the following runtime numbers:

AVX512:

Parameters                                                     | Runtime without patch (ns) | Runtime with patch (ns) | Speedup
---------------------------------------------------------------|----------------------------|-------------------------|---------
BM_RHS_NAME(PackRhs, 128, 24, 24, 3, 64, 5, 5, 1, 1, 256, 56)  |  41350                     | 15073                   | 2.74X
BM_RHS_NAME(PackRhs, 32, 64, 64, 32, 64, 5, 5, 1, 1, 256, 56)  |   7277                     |  7341                   | 0.99X
BM_RHS_NAME(PackRhs, 32, 64, 64, 32, 64, 5, 5, 2, 2, 256, 56)  |   8675                     |  8681                   | 1.00X
BM_RHS_NAME(PackRhs, 32, 64, 64, 30, 64, 5, 5, 1, 1, 256, 56)  |  24155                     | 16079                   | 1.50X
BM_RHS_NAME(PackRhs, 32, 64, 64, 30, 64, 5, 5, 2, 2, 256, 56)  |  25052                     | 17152                   | 1.46X
BM_RHS_NAME(PackRhs, 32, 256, 256, 4, 16, 8, 8, 1, 1, 256, 56) |  18269                     | 18345                   | 1.00X
BM_RHS_NAME(PackRhs, 32, 256, 256, 4, 16, 8, 8, 2, 4, 256, 56) |  19468                     | 19872                   | 0.98X
BM_RHS_NAME(PackRhs, 32, 64, 64, 4, 16, 3, 3, 1, 1, 36, 432)   | 156060                     | 42432                   | 3.68X
BM_RHS_NAME(PackRhs, 32, 64, 64, 4, 16, 3, 3, 2, 2, 36, 432)   | 132701                     | 36944                   | 3.59X

AVX2:

Parameters                                                     | Runtime without patch (ns) | Runtime with patch (ns) | Speedup
---------------------------------------------------------------|----------------------------|-------------------------|---------
BM_RHS_NAME(PackRhs, 128, 24, 24, 3, 64, 5, 5, 1, 1, 256, 56)  | 26233                      | 12393                   | 2.12X
BM_RHS_NAME(PackRhs, 32, 64, 64, 32, 64, 5, 5, 1, 1, 256, 56)  |  6091                      |  6062                   | 1.00X
BM_RHS_NAME(PackRhs, 32, 64, 64, 32, 64, 5, 5, 2, 2, 256, 56)  |  7427                      |  7408                   | 1.00X
BM_RHS_NAME(PackRhs, 32, 64, 64, 30, 64, 5, 5, 1, 1, 256, 56)  | 23453                      | 20826                   | 1.13X
BM_RHS_NAME(PackRhs, 32, 64, 64, 30, 64, 5, 5, 2, 2, 256, 56)  | 23167                      | 22091                   | 1.09X
BM_RHS_NAME(PackRhs, 32, 256, 256, 4, 16, 8, 8, 1, 1, 256, 56) | 23422                      | 23682                   | 0.99X
BM_RHS_NAME(PackRhs, 32, 256, 256, 4, 16, 8, 8, 2, 4, 256, 56) | 23165                      | 23663                   | 0.98X
BM_RHS_NAME(PackRhs, 32, 64, 64, 4, 16, 3, 3, 1, 1, 36, 432)   | 72689                      | 44969                   | 1.62X
BM_RHS_NAME(PackRhs, 32, 64, 64, 4, 16, 3, 3, 2, 2, 36, 432)   | 61732                      | 39779                   | 1.55X

All benchmarks on Intel Skylake server with 8 cores.
 2019-04-20 06:46:43 +00:00
Gael Guennebaud
eb4c6bb22d Fix conflicts and merge 2019-01-30 15:57:08 +01:00
Rasmus Munk Larsen
7401e2541d Fix compilation error for logical packet ops with older compilers. 2019-01-16 14:43:33 -08:00
Rasmus Munk Larsen
fcfced13ed Rename pones -> ptrue. Use _CMP_TRUE_UQ where appropriate. 2019-01-09 17:20:33 -08:00
Rasmus Munk Larsen
8f04442526 Collapsed revision 
* Collapsed revision
* Add packet up "pones". Write pnot(a) as pxor(pones(a), a).
* Collapsed revision
* Simplify a bit.
* Undo useless diffs.
* Fix typo.
 2019-01-09 16:34:23 -08:00
Rasmus Munk Larsen
e00521b514 Undo useless diffs. 2019-01-09 16:32:53 -08:00
Rasmus Munk Larsen
f2767112c8 Simplify a bit. 2019-01-09 16:29:18 -08:00
Rasmus Munk Larsen
cb955df9a6 Add packet up "pones". Write pnot(a) as pxor(pones(a), a). 2019-01-09 16:17:08 -08:00
Rasmus Munk Larsen
055f0b73db Add support for pcmp_eq and pnot, including for complex types. 2019-01-07 16:53:36 -08:00
Gustavo Lima Chaves
e763fcd09e Introducing "vectorized" byte on unpacket_traits structs 
This is a preparation to a change on gebp_traits, where a new template
argument will be introduced to dictate the packet size, so it won't be
bound to the current/max packet size only anymore.

By having packet types defined early on gebp_traits, one has now to
act on packet types, not scalars anymore, for the enum values defined
on that class. One approach for reaching the vectorizable/size
properties one needs there could be getting the packet's scalar again
with unpacket_traits<>, then the size/Vectorizable enum entries from
packet_traits<>. It turns out guards like "#ifndef
EIGEN_VECTORIZE_AVX512" at AVX/PacketMath.h will hide smaller packet
variations of packet_traits<> for some types (and it makes sense to
keep that). In other words, one can't go back to the scalar and create
a new PacketType, as this will always lead to the maximum packet type
for the architecture.

The less costly/invasive solution for that, thus, is to add the
vectorizable info on every unpacket_traits struct as well.
 2018-12-19 14:24:44 -08:00
Gael Guennebaud
81c27325ae bug #1641: fix testing of pandnot and fix pandnot for complex on SSE/AVX/AVX512 2018-12-08 14:27:48 +01:00
Gael Guennebaud
7134fa7a2e Fix compilation with MSVC by reverting to char* for _mm_prefetch except for PGI (the later being the one that has the wrong prototype). 2018-06-07 09:33:10 +02:00
Gael Guennebaud
49262dfee6 Fix compilation and SSE support with PGI compiler 2018-05-29 15:09:31 +02:00
Gael Guennebaud
b240080e64 bug #1436: fix compilation of Jacobi rotations with ARM NEON, some specializations of internal::conj_helper were missing. 2017-06-15 10:16:30 +02:00
Gael Guennebaud
598de8b193 Add pinsertfirst function and implement pinsertlast for complex on SSE/AVX. 2016-11-02 10:38:13 +01:00
Gael Guennebaud
70404e07c2 Workaround clang -Wdocumentation warning about "/*<" 2015-12-30 16:46:45 +01:00
Gael Guennebaud
2c676ddb40 Handle various TODOs in SSE vectorization (remove splitted storeu, enable SSE3 integer vectorization, plus minor tweaks) 2015-10-06 15:43:27 +02:00
Gael Guennebaud
ce57dbd937 Let unpacket_traits<> exposes the required alignment and make use of it everywhere 2015-08-07 10:44:01 +02:00
Gael Guennebaud
91ab2489dd Fix compilation with GCC/AVX (workaround __m128 and __m256 being the same type with default ABI) 2015-02-17 16:08:07 +01:00
Gael Guennebaud
45cbb0bbb1 The usage of DenseIndex is deprecated, so let's replace DenseIndex by Index 2015-02-16 15:05:41 +01:00
Benoit Steiner
16047c8d4a Pulled in the latest changes from the Eigen trunk 2014-08-13 22:25:29 -07:00
Gael Guennebaud
b47ef1431f Fix many long to int implicit conversions 2014-07-08 16:47:11 +02:00
Benoit Steiner
29aebf96e6 Created the pblend packet primitive and implemented it using SSE and AVX instructions. 2014-06-06 20:18:44 -07:00
Gael Guennebaud
3d8d0f6269 Enable vectorization of pack_rhs with a column-major RHS. 
Rename and generalize Kernel<*> to PacketBlock<*,N>.
 2014-04-25 10:56:18 +02:00
Gael Guennebaud
5c5231ab71 Workaround gcc's default ABI not being able to distinghish between vector types of different sizes. 2014-04-22 16:03:19 +02:00
Gael Guennebaud
d5a795f673 New gebp kernel handling up to 3 packets x 4 register-level blocks. Huge speeup on Haswell. 
This changeset also introduce new vector functions: ploadquad and predux4.
 2014-04-16 17:05:11 +02:00
Gael Guennebaud
1c0728043a Workaround alignment warnings 2014-03-30 22:43:47 +02:00
Gael Guennebaud
10aa14592a Add a mechanism to recursively access to half-size packet types 2014-03-28 10:18:04 +01:00
Benoit Steiner
8a94cb3edd Implemented the SSE version of the gather and scatter packet primitives. 2014-03-27 18:29:01 -07:00
Gael Guennebaud
052aedd394 Implement pcplflip, palign, predux and the likes from AVC/complexes 2014-03-27 14:47:00 +01:00
Benoit Steiner
a419cea4a0 Created the ptranspose packet primitive that can transpose an array of N packets, where N is the number of words in each packet. This primitive will be used to complete the vectorization of the gemm_pack_lhs and gemm_pack_rhs functions. 
Implemented the primitive using SSE instructions.
 2014-03-26 19:03:07 -07:00
Gael Guennebaud
62670c83a0 Fix bug #314: move remaining math functions from internal to numext namespace 2013-06-10 23:40:56 +02:00
Gael Guennebaud
e4ec63aee7 Suppress annoying "may be used uninitialized in this function" warning with gcc >= 4.6 2013-01-24 11:59:17 +01:00
Benoit Jacob
69124cfca2 Automatic relicensing to MPL2 using Keirs script. Manual fixup follows. 2012-07-13 14:42:47 -04:00
Jitse Niesen
3c412183b2 Get rid of include directives inside namespace blocks (bug #339). 2012-04-15 11:06:28 +01:00
Gael Guennebaud
634fedaf68 proper C++ casting 2012-01-31 18:56:25 +01:00
Gael Guennebaud
9c86ee2695 fix static inline versus inline static issues (the former is the correct order) 2012-01-31 12:58:52 +01:00
Gael Guennebaud
c121e6f390 implement ploaddup for complex and SSE/NEON even though they are not used in practice 2011-02-23 16:31:42 +01:00
Gael Guennebaud
6e01780541 fix a couple of issues with pcplxflip 2011-02-23 17:51:40 +03:00
Gael Guennebaud
fb1a29fed5 fix ICE and warning with gcc 4.2.4 2011-02-21 16:11:18 +01:00
Gael Guennebaud
8f8c67b8bd fix bug #186 (in 32 bits mode, gcc 4.3 messed up with pfirst for complex<float>) 2011-02-18 15:47:17 +01:00
Gael Guennebaud
d6c4ca4845 fix redundancy 2011-02-09 13:44:05 +01:00
Gael Guennebaud
c0d5131435 workaround gcc 4.2.1 ICE (fix bug #145) 2011-02-09 13:04:35 +01:00
Gael Guennebaud
5887a086cf fix SSE3 issue (infinite loop after the ei_ => internal change) - this fix bug #174 2011-02-03 17:55:24 +01:00
Gael Guennebaud
da05b6af0e fix some remainign issue with ei_ -> internal change 2010-11-16 15:54:48 +01:00
Benoit Jacob
4716040703 bug #86 : use internal:: namespace instead of ei_ prefix 2010-10-25 10:15:22 -04:00
Gael Guennebaud
d4b664c4cd fix ugly conversion from double[2] to complex 2010-08-19 14:47:58 +02:00
Gael Guennebaud
c7f40e522e merge 2010-07-22 13:21:06 +02:00
Gael Guennebaud
0dfc5b296b fix strict aliasing issue 2010-07-22 13:16:53 +02:00