GitHub-Proxy/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2025-05-04 01:34:07 +08:00
Code Issues Packages Projects Releases Wiki Activity
9,356 Commits 12 Branches 89 Tags
Commit Graph

55 Commits

Author SHA1 Message Date
Hugh Perkins
9341f258d4 Add labels to #ifdef, in TensorReductionCuda.h 2017-06-06 15:51:06 +01:00
Benoit Steiner
a71943b9a4 Made the Tensor code compile with clang 3.9 2017-03-02 10:47:29 -08:00
Igor Babuschkin
18c67df31c Fix remaining CUDA >= 300 checks 2016-08-18 17:18:30 +01:00
Igor Babuschkin
1569a7d7ab Add the necessary CUDA >= 300 checks back 2016-08-18 17:15:12 +01:00
Igor Babuschkin
841e075154 Remove CUDA >= 300 checks and enable outer reductin for doubles 2016-08-06 18:07:50 +01:00
Igor Babuschkin
9537e8b118 Make use of atomicExch for atomicExchCustom 2016-08-05 14:29:58 +01:00
Igor Babuschkin
eeb0d880ee Enable efficient Tensor reduction for doubles 2016-07-01 19:08:26 +01:00
Benoit Steiner
cb2d8b8fa6 Made it possible to compile reductions for an old cuda architecture and run them on a recent gpu. 2016-06-29 15:42:01 -07:00
Benoit Steiner
b2a47641ce Made the code compile when using CUDA architecture < 300 2016-06-29 15:32:47 -07:00
Benoit Steiner
37638dafd7 Simplified the code that dispatches vectorized reductions on GPU 2016-06-09 10:29:52 -07:00
Benoit Steiner
aa33446dac Improved support for vectorization of 16-bit floats 2016-06-09 08:22:27 -07:00
Benoit Steiner
7ef9f47b58 Misc small improvements to the reduction code. 2016-06-06 14:09:46 -07:00
Benoit Steiner
c2a102345f Improved the performance of full reductions. 
AFTER:
BM_fullReduction/10        4541       4543     154017  21.0M items/s
BM_fullReduction/64        5191       5193     100000  752.5M items/s
BM_fullReduction/512       9588       9588      71361  25.5G items/s
BM_fullReduction/4k      244314     244281       2863  64.0G items/s
BM_fullReduction/5k      359382     359363       1946  64.8G items/s

BEFORE:
BM_fullReduction/10        9085       9087      74395  10.5M items/s
BM_fullReduction/64        9478       9478      72014  412.1M items/s
BM_fullReduction/512      14643      14646      46902  16.7G items/s
BM_fullReduction/4k      260338     260384       2678  60.0G items/s
BM_fullReduction/5k      385076     385178       1818  60.5G items/s
 2016-06-03 17:27:08 -07:00
Benoit Steiner
873e6ac54b Silenced compilation warning generated by nvcc. 2016-06-01 14:20:50 -07:00
Benoit Steiner
d27b0ad4c8 Added support for mean reductions on fp16 2016-06-01 11:12:07 -07:00
Benoit Steiner
5aeb3687c4 Only enable optimized reductions of fp16 if the reduction functor supports them 2016-05-31 10:33:40 -07:00
Benoit Steiner
36369ab63c Resolved merge conflicts 2016-05-26 13:39:39 -07:00
Benoit Steiner
28fcb5ca2a Merged latest reduction improvements 2016-05-26 12:19:33 -07:00
Benoit Steiner
c1c7f06c35 Improved the performance of inner reductions. 2016-05-26 11:53:59 -07:00
Benoit Steiner
0835667329 There is no need to make the fp16 full reduction kernel a static function. 2016-05-24 23:11:56 -07:00
Benoit Steiner
8d06c02ffd Allow vectorized padding on GPU. This helps speed things up a little. 
Before:
BM_padding/10            5000000        460   217.03 MFlops/s
BM_padding/80            5000000        460 13899.40 MFlops/s
BM_padding/640           5000000        461 888421.17 MFlops/s
BM_padding/4K            5000000        460 54316322.55 MFlops/s
After:
BM_padding/10            5000000        454   220.20 MFlops/s
BM_padding/80            5000000        455 14039.86 MFlops/s
BM_padding/640           5000000        452 904968.83 MFlops/s
BM_padding/4K            5000000        411 60750049.21 MFlops/s
 2016-05-17 09:13:27 -07:00
Benoit Steiner
83dfb40f66 Turnon the new thread pool by default since it scales much better over multiple cores. It is still possible to revert to the old thread pool by compiling with the EIGEN_USE_SIMPLE_THREAD_POOL define. 2016-05-13 17:23:15 -07:00
Benoit Steiner
c4fc8b70ec Removed unnecessary thread synchronization 2016-05-13 10:49:38 -07:00
Benoit Steiner
217d984abc Fixed a typo in my previous commit 2016-05-11 10:22:15 -07:00
Benoit Steiner
08348b4e48 Fix potential race condition in the CUDA reduction code. 2016-05-11 10:08:51 -07:00
Benoit Steiner
4ede059de1 Properly gate the use of half2. 2016-05-10 17:04:01 -07:00
Benoit Steiner
0eb69b7552 Small improvement to the full reduction of fp16 2016-05-10 11:58:18 -07:00
Benoit Steiner
4013b8feca Simplified the reduction code a little. 2016-05-10 09:40:42 -07:00
Benoit Steiner
4670d7d5ce Improved the performance of full reductions on GPU: 
Before:
BM_fullReduction/10       200000      11751     8.51 MFlops/s
BM_fullReduction/80         5000     523385    12.23 MFlops/s
BM_fullReduction/640          50   36179326    11.32 MFlops/s
BM_fullReduction/4K            1 2173517195    11.50 MFlops/s

After:
BM_fullReduction/10       500000       5987    16.70 MFlops/s
BM_fullReduction/80       200000      10636   601.73 MFlops/s
BM_fullReduction/640       50000      58428  7010.31 MFlops/s
BM_fullReduction/4K         1000    2006106 12461.95 MFlops/s
 2016-05-09 17:09:54 -07:00
Benoit Steiner
2dde1b1028 Don't crash when attempting to reduce empty tensors. 2016-04-20 18:08:20 -07:00
Benoit Steiner
5b1106c56b Fixed a compilation error with nvcc 7. 2016-04-19 14:57:57 -07:00
Benoit Steiner
7129d998db Simplified the code that launches cuda kernels. 2016-04-19 14:55:21 -07:00
Benoit Steiner
884c075058 Use numext::ceil instead of std::ceil 2016-04-19 14:33:30 -07:00
Benoit Steiner
edc679f6c6 Fixed compilation warning 2016-03-18 07:12:34 -07:00
Benoit Steiner
68ac5c1738 Improved the performance of large outer reductions on cuda 2016-02-29 18:11:58 -08:00
Benoit Steiner
b2075cb7a2 Made the signature of the inner and outer reducers consistent 2016-02-29 10:53:38 -08:00
Benoit Steiner
3284842045 Optimized the performance of narrow reductions on CUDA devices 2016-02-29 10:48:16 -08:00
Benoit Steiner
34057cff23 Fixed a race condition that could affect some reductions on CUDA devices. 2016-01-15 15:11:56 -08:00
Benoit Steiner
aed4cb1269 Use warp shuffles instead of shared memory access to speedup the inner reduction kernel. 2016-01-14 21:45:14 -08:00
Benoit Steiner
8fe2532e70 Fixed a boundary condition bug in the outer reduction kernel 2016-01-14 09:29:48 -08:00
Benoit Steiner
c5e6900400 Silenced a few compilation warnings. 2016-01-11 17:06:39 -08:00
Benoit Steiner
01c55d37e6 Deleted unused variable. 2016-01-11 15:53:19 -08:00
Benoit Steiner
0504c56ea7 Silenced a nvcc compilation warning 2016-01-11 15:49:21 -08:00
Benoit Steiner
b523771a24 Silenced several compilation warnings triggered by nvcc. 2016-01-11 14:25:43 -08:00
Benoit Steiner
2c3b13eded Merged in jeremy_barnes/eigen/shader-model-3.0 (pull request PR-152) 
Alternative way of forcing instantiation of device kernels without causing warnings or requiring device to device kernel invocations.
 2016-01-11 11:43:37 -08:00
Benoit Steiner
780623261e Re-enabled the optimized reduction CUDA code. 2016-01-11 09:07:14 -08:00
Jeremy Barnes
403a7cb6c3 Alternative way of forcing instantiation of device kernels without 
causing warnings or requiring device to device kernel invocations.

This allows Tensorflow to work on SM 3.0 (ie, Amazon EC2) machines.
 2016-01-10 22:39:13 -05:00
Benoit Steiner
53749ff415 Prevent nvcc from miscompiling the cuda metakernel. Unfortunately this reintroduces some compulation warnings but it's much better than having to deal with random assertion failures. 2016-01-08 13:53:40 -08:00
Benoit Steiner
cfff40b1d4 Improved the performance of reductions on CUDA devices 2016-01-04 17:25:00 -08:00
Benoit Steiner
a1e08fb2a5 Optimized the configuration of the outer reduction cuda kernel 2015-12-22 16:30:10 -08:00