aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2018-11-02COMPMID-1413 - Improve the performance of GEMMLowp with 8 bit dot product on ↵Gian Marco Iodice
OpenCL COMPMID-1424 - Add dot product support for CLDepthwise QASYMM8 3x3 NHWC non-unit stride With this patch we are able to improve the performance of MobileNet v1-qasymm8 by 37 % Tried to use the dot product instruction in CLDepthwise QASYMM8 3x3 NHWC non-unit stride but I have not seen any benefit (maybe because we have few arithemtic operation and we do not have more load instructions). However Depthwise convolution has been improved by 30% Change-Id: Id768a99c2e53a04276707e427af5d0ec93419ada Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155082 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1029: Collapse CLWinogradInputTransform/CLWinogradOutputTransformGeorgios Pinitas
Change-Id: I051748502ca24b9952e7313524bbfd708162efb4 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155166 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1695: (Nightly) bad_alloc in CLIm2ColGeorgios Pinitas
Clear CLContext in a more regular basis to make the driver release memory back to the system. Change-Id: I0df847766f57719433bbaeada45fe630e38c9541 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155435 Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1696: (Nighlty) CLDepthwiseConvolution FP16 mismatchesGeorgios Pinitas
Increases relative tolerance slightly as error was quite small. Change-Id: I4789c5e3eeb4f2d3aaf2b4c76966474f045af4c1 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155418 Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451: Fix CL/NEPermuteKernel PermuteVection checkIsabella Gottardi
COMPMID-1690: Add tests for NEPermute with PermutationVector dimension > 3 Change-Id: I4bfc6ff88cd46863c2e39975b5663c624db1a63d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155316 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1685: Allow build_dir to be an absolute pathAnthony Barbier
Change-Id: I875ffe0ccec3aa4f53bfb68d82e2a7292ab83358 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155348 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451: Fix validation issue in CLReduceMeanMichalis Spyrou
Change-Id: Ie1bcdd9dca2dc3b26003790a19cc80bb953385b2 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155373 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451: FP16 L2Normalization failuresMichalis Spyrou
Set input range to [-1, 1] in order to avoid inf values when calculating sqrt. Change-Id: I18f1e427baa7830fdc587bedf27a92d78c72f49b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155397 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-711: add GEMMInterleave4X4 benchmarking tests.Pablo Tello
Change-Id: I7a32becd78fc231d11d50c6ff58892f4acb0ccda Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155224 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1451: Fix inlines in cl helpersGeorgios Pinitas
Change-Id: I9cb725a8052091469904ecc7cfffa4add9914ffb Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155261 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Manuel Bottini <manuel.bottini@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1530 error: dereferencing type-punned pointer will break ↵Michalis Spyrou
strict-aliasing rules Change-Id: I9e54d07cf1d77c14f124056d3724b49981bf3f97 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155292 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451: Remove invalid bias check in NEDeconvolutionLayerGeorgios Pinitas
Change-Id: Ibb6b0ceed19111c01fcc96eb50e461f9811b61b9 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155260 Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451: DATA_UPDATE Update data folderMichalis Spyrou
Change-Id: I186b8875089ff77768db051f58f2968e3e9505e3 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155202 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1680: (Nighlty) CLBBoxTransform mismatchesgiuros01
Instead of changing the tolerances I increased the sizes of the input. In this way, for a single mismatch, as it was the case, we are below the 1% tolerance set. Change-Id: I787261a1d1adb559c1687b7bd1e0317a72594130 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155168 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1682 Nightly NEON Reduction operation mismatchesMichalis Spyrou
Increase tolerance for fp32 and added absolute tolerance Change-Id: Iff828457b514d6301ed2c8e04b66ce86867f72b6 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155086 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com>
2018-11-02COMPMID-1681: (Nightly) NEWidthConcatenateLayer failsMichele Di Giorgio
NEWidthConcatenateLayerKernel works with 4D tensors too, hence the check has been removed and tests have been added. Change-Id: I73814cabe5fae975a44cc1a03b092c552497e57d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155070 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com>
2018-11-02COMPMID-1451 Properly remove dimensions in CLReduceMeanMichalis Spyrou
Change-Id: I7bd4a8ce81483ba56686b765ca3caabebe42882d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155000 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
2018-11-02COMPMID-1673: Collapse window in CLArithmeticAddition when one operand is a ↵Michele Di Giorgio
vector When one of the operands is a vector, the kernel does a broadcast addition and the window is not collapsed. This represent an issue because it leads to a lot of enqueues that increases the time taken by the OpenCL driver. This patch allows to collapse the window when one of the two operands is a vector. Furthermore, it adds LWS tuner to the kernel. It also changes the number of elements processed per iteration to 8 to make better usage of the cache. Change-Id: I5f09ab0ddcffb3b7f9326a987c79a997b2d7fa8c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155003 Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1641: Create Facebook's MaskRCNN v2 graph example (3RDPARTY_UPDATE)Georgios Pinitas
Change-Id: Id106d53b9477298a117a5195f3fc5b0f36003c35 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/131903 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1451: Perform CLOutputStage using floats.Georgios Pinitas
Change-Id: Ic8312a5b6790aa7cd4468d42f08d557ad40e9441 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154570 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1451: Fuse activation in DepthwiseConvolution.Georgios Pinitas
Change-Id: Id964d9068e18aaa13ab8adcbf7a9375b034ea6c3 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154651 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1327: Add support for BBoxTransform operator in CLgiuros01
Change-Id: I91865506166951b3bf7f06a0b2d4cde925cfefb6 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/153447 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1451:Fuse RELU,LU_BOUNDED_RELU with requantization in ↵Georgios Pinitas
NEGEMMConvolutionLayer. Change-Id: Iea5f2c5bcac8051c4c7655a6eabb2c43772eb31f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154104 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1632 Add CLL2NormalizationLayer for NHWC and FP32Michalis Spyrou
Change-Id: Iae22554d5fe893fd22a000eab5bfd8275ea06eb3 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154102 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1523: Fuse BN node with convolution.Georgios Pinitas
Change-Id: I146936c9e98b343496a4b61cdbadf0eaa38e885a Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154008 Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1667: Add 4D tensors support to CLWidthConcatenateLayerKernelMichele Di Giorgio
Change-Id: Ibc0b1242804c2fdb183825406e3c78bd0d1d3564 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154368 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1580 Implement ReduceMean in NEONMichalis Spyrou
Change-Id: Id974efad304c2513b8824a6561ad45ee60b9e7fb Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/153763 Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com> Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1659: (Nightly) CLGEMMConvolutionLayer QASYMM8 TensorShape errorIsabella Gottardi
Change-Id: Ib4ca28b82bd82f0ed4d2c906185d3f4010246616 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/153986 Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451 - Enabled NHWC in AlexNet, InceptionV3 and InceptionV4Gian Marco Iodice
Enabled NHWC as default data layout in AlexNet, InceptionV3 and InceptionV4 on NEON Change-Id: I205ad7a1f5dbf482340182de4fac8f3a24ef2705 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154141 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1661 Nightly - CLReductionOperation FP16 failuresMichalis Spyrou
Change-Id: If496709958bf29589601eac62a268819736a4fd2 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/154173 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451: Fix allocation of weights in DeconvolutionMichele Di Giorgio
Change-Id: If3ca0b034a7448df1e5349b51a2b124f1b4e99c1 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/153956 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1586: Add support for NHWC CLDeconvolutionLayerMichele Di Giorgio
COMPMID-1651: Fix QASYMM8 CLDeconvolutionLayer This patch also extends the range of values used for testing Convolution and Deconvolution to cover quantized [-1.0f, 1.0f]. Change-Id: I8b280669db67bb3ec25bf5d411c8f5954f5b0dab Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/149869 Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1574 Implement ReduceMean in OpenCLMichalis Spyrou
Change-Id: Id331199f569f52a37280a9ada5bf84694580b93c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/152843 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
2018-11-02COMPMID-1657 - Fix CPU info in the json file outputGian Marco Iodice
Change-Id: I05a1b871746a32ccc1c3ecec97b8266767c9d0a7 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/153715 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1451: Reverting changes for CLGEMM and CLGEMMLowp previuosly done ↵Isabella Gottardi
(384496) Mirroring CLGEMM behaviour to CLGEMMLowp Change-Id: I308b54e2c0de131a5322b77e83e7454db498d692 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/153175 Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1631: Fixed test framework memory corruption.Pablo Tello
Change-Id: I2c4dcedcd3b56e41174eebbbacd47be4e968d34d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/152767 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Vidhya Sudhan Loganathan <vidhyasudhan.loganathan@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1451: Fix NormalizationLayer accross width normalization.Georgios Pinitas
NEON and CL normalization layer was generating invalida results for radius > 4. Change-Id: I15d846405e6b3492fe44920bbf8cadceb4e5258f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/153161 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Matteo Martincigh <matteo.martincigh@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1621 Deconvolution wrong output calculationMichalis Spyrou
Change-Id: Ida71312bcf6dbd854f2ab1efc65f74910c79e152 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/151510 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
2018-11-02COMPMID-1451: Fix compilation issues under gcc 8Georgios Pinitas
Change-Id: I05d3447336ee0bf330e2a0c58fc6904be1db8f83 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/152626 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1628: NEWinograd mismatches in the nightly.Pablo Tello
Since this is floating point arithmetic the Winograd results will not be exactly the same as direct convolution. Changed to use relative tolerance for the nightly tests. Change-Id: I45c6d60a097c2d4fb53650a2a33eb29a3e51d7ec Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/152324 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
2018-11-02COMPMID-1623: NEWinograd reduce the number of output tiles.Pablo Tello
Change-Id: I4d9240924fe483d2dd127ad6a4ae6f8066f61bd1 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/151893 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Andrew Mundy <andrew.mundy@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1622 : (Nightly) Seg fault on androidVidhya Sudhan Loganathan
call to print_cpu_info moved to main Change-Id: I6d82649964542df4e944bc79e4c16f0813976295 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/152695 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1451: Enable dot kernels in NEGEMMAssembly functionsGeorgios Pinitas
Change-Id: I9dd26b80025ea3a4c66f5f0bf41b7a98dd0d3aa4 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/152549 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1630: Fixed broken DOTPROD detection.Pablo Tello
Change-Id: Ic857b938130c3d8713c9b414249cfabdcf846c2b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/152451 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Vidhya Sudhan Loganathan <vidhyasudhan.loganathan@arm.com>
2018-11-02COMPMID-1622 : (Nightly) Seg fault on androidVidhya Sudhan Loganathan
conditionally compile the std::cout that was causing the fault Change-Id: I7f50151ab88f19ed6eec1be11ca975614653e359 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/151762 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1607 - (Nightly) CLGEMMLowpMatrixMultiplyCore errors and mismatchesIsabella Gottardi
Change-Id: I5f2e6843526cb154176a5b113627d4f36c3a8edd Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/150967 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1627: (nightly) ROIAlign failuregiuros01
Change-Id: I062e7673f26d5267ed113eae7edd361d05d6de73 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/151968 Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1546 Optimize PoolingLayer NHWC on NEON for all data typesMichalis Spyrou
Change-Id: I4920e43059a713126f15493f38fe50f07d0a8c7f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/151087 Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1610: Fixed CLDirectConvolution mismatchesPablo Tello
Kernel size 5x5 layout NHWC. Change-Id: Ia82ff211d1c954df228962b5c2c5ad8df7112449 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/151740 Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
2018-11-02COMPMID-1618 : Mobilenet QASYMM8 does not use dot prod kernel when ↵Vidhya Sudhan Loganathan
instrumentation timer is enabled issue description: cpu info was neither initialised in scheduler timer interceptor nor was it retrieved from real scheduler. Change-Id: Ibe75b0704250d99682ed866deaf3ddf5fda22de5 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/151451 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com>
generated by cgit v1.2.1 (git 2.23.0) at 2024-06-25 22:32:41 +0000