Age | Commit message (Collapse) | Author |
|
This patch makes col2im on OpenCL 2 times faster
Change-Id: I8d90f5a72a050355ca1fd13433d8c2c26e5e33f5
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119442
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Changes and simplifies the validation to divide the scale in integer format instead of double.
Change-Id: Ib9156e9515e4e542391eeda11548f3d15613a0af
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119256
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
|
|
Change-Id: Ib8100e7c659c49694c746fa3f36ce20f44f6929f
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117804
Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Number of threads in the scheduler were set after example configuration
leading to memory corruption in case more threads were used.
Change-Id: I221a026196cd64f1805d31596f0488a247a4bfab
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119196
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com>
|
|
This patch brings the MACs utilisation up to 25 % when both stride_x and stride_y are equal to 1
Performance reported in the following confluence page:
https://confluence.arm.com/display/MLENG/Depthwise+convolution+3x3+FP32+performance%3A+ACL+18.02
Change-Id: Ida1b64be9a88805902a3d90194559b58eb1224a3
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119068
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Removed double managment of the same tensor object
Change-Id: Ibc74cd8c7bd199cd473ff68f692840cbf01b27b3
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119119
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com>
|
|
The LWS hint has been applied for optimized cases 1x1 and 3x3
Change-Id: I6b4bfe2f9f7da627052336889b8a18d279fe2675
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119162
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
It's not safe to accumulate two u8xu8 results into a u16 accumulator.
This changes the kernel to use uadalp after every single multiply.
Correct the test fixture as well.
Change-Id: I011b90033c4673e55b843d079e3f7d185b1df330
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119096
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I5a420da6a8041f9ff6d0811815f2fc74c85c56a8
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119014
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: I84a914c13b162c4f74321c9cafc30a18ad4ebbdb
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118797
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Introduced optimizations for 1x1, 3x3, 5x5 and 11x11
Change-Id: Ibb7f7a9fbec01a7684746ed8513634078126e452
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118107
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
|
|
Change-Id: If6f3888a035b557a6c369efa22b56d6c8d3efbd3
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118789
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
|
|
example
Change-Id: Ic639d51fb5dd4f78912a9b11abc7df79d205a22b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118843
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I12d4af007c123b19925ceb5e3c84285e096bc13b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118718
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: Ia30ec2afce0aafcd39f41440efb972b18bbda9f8
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118657
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Change-Id: I5296815cf04e5f805d6523196567b6c01715c8b5
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118711
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: Iebd2a8fece1af87c93d6795e176d8c37ca64bbf6
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118187
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I71f67789648ef05ccdedce77c7427bc0127b3a69
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/116741
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: Ie4ac7f61675c1fb9b1748d6784fccb26f058832a
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118635
Reviewed-by: Robert Hughes <robert.hughes@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I83d0f2bc8e0ebfdc0b60931f2c5acf0469caf886
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118696
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Arm Cortex-A55 FPGA with 8 CPUs with lots of flags).
Change-Id: I493fb1013c6c25d9b9c809705b1ee24abac1d8d1
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118456
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
1) Removed the example files winograd_layer.hpp/cpp
2) Teplatized winograd transform kernels
Change-Id: I7045fa0b801b9d30a11275914aaa2dafd254aed2
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118332
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I14ff5e2964328d22c0bba5a77683e07f0c7920e9
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118389
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
issue on OpenGL ES
Change-Id: I7a8489bb0fddc72899ea165e414ee87bdbfb45b3
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118106
Reviewed-by: Joel Liang <joel.liang@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I5366d11aefdb8f3ba7326ed7527eb216c4de0668
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118372
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: Ie480332e6e302edd406627e90be0d7df3e61dde5
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118303
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
investigated
Change-Id: I5a69198bfd60d9cdd061f2db9838d9f0df9ecc23
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118454
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Also, added instrumentation to support generic tensor broadcasting for
NEON and CL backends.
Change-Id: I1bc5747a286e1a4b464c209067581e103d473b9a
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/114201
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: Icbb43de7642e2b433d7471d70b9dbbde850989d3
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118197
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Change-Id: I2d3cc9668852a1ba414fc3148866df408f770dc8
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118308
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
specific dataset
Change-Id: I227e90445715c3bd394e49930b010c0a5f5ca177
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118108
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Joel Liang <joel.liang@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: Ic1f215c1ae85ad5c516cc3600447a50bba77ebc1
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117668
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I38ae204632ae27c5fe7a0131462343397899868c
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118120
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Fully Connected test names are not unique
Change-Id: Ie4654cc1cb4720c51a3114162043562d5cbc6d28
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118126
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: I880ac3a1c3f5ea09ccefe27d9ee40bd60afcea2b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118056
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: I0a7ea4cde1dbf8edd28908dfff80928ef7e996c4
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117647
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: Iff50adf2993bd69c2696a47559d6b2e0011fed87
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110177
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I80437f7ba6e4b8ec1fb145300a017b3688f3f2b6
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118086
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Some minor improvements in the test fixture, for example making sure
the values in the mapx and mapy tensors are in the range of [-5, in_width+5]
and [-5,in_height].
Tolerance was changed to 0, no mismatches expected.
Change-Id: I2fad06defb293bf9fdd1988799b19547c102dee5
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118044
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: Ic460695b8a203c1080ea177b5463b48b07b70c4b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118075
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Joel Liang <joel.liang@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I4f2cca52caf210fdb7d6bb7e9436ac51cb5088b4
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/112398
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Change-Id: Ia2874d30780cb597a6e5039120815f2368911e0c
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118024
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
1) Updated to the latest code from the RSH repo.
2) Moved winograd transforms into kernels.
3) Added support for biases
Change-Id: I7f39f34a599b49d7d9b549cc10a4f4d4a8007ab8
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117474
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I33cf54e68f6c097ac58b6f16c3f9a720978f09cd
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117289
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Currently we output an array of timestamps: queued, submitted, start, end
This patch instead only output end-start (i.e the time it took to execute the kernel on the GPU)
Change-Id: Ic3c2b68128f6acd6bb018b7b3ead0b69dd5aca59
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117865
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Kevin Petit <kevin.petit@arm.com>
|
|
Change-Id: Iec82a91ad351cfe8d07d0976a24bd42f4703177a
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/116833
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
|
|
Change-Id: If8c1e0103ae2e3dfde3d0b9f23575c0e904c7f30
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117961
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Refactored the console printer too (So that we can re-use the code if needed)
Change-Id: I16a0f70104f82f07cd59900b383038fa5a76e1bc
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117858
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Changed CLReductionOperationKernel: Now each kernel computes
a 2D slice instead of 1D. This reduces the memory footprint
from around 1.6Gb for a 4k input image to a few Mb, which was
caused by the __local memory and was probably the cause for this bug.
Change-Id: I71ac71ff09b041c945a134177600f0f3475e48cf
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117835
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
it supports asymmetric padding
Add asymmetric padding support for NEPoolingLayer
Change-Id: Ia5cc660aeca636c3c45df4916a28974cc2b7f2f4
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117275
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|