commit | 73ccfb4e007b67efde34348ced025eb75f9431f2 | [log] [tgz] |
---|---|---|
author | Frank Barchard <fbarchard@google.com> | Wed Dec 11 22:15:22 2019 -0800 |
committer | XNNPACK Team <xnnpack-github-robot@google.com> | Wed Dec 11 22:27:01 2019 -0800 |
tree | 710f27731b9f15e9ca66903cf9ccc778676f5e0b | |
parent | a84e40bd4ca96343d4db4c533f5c4f66a31ab31b [diff] |
Move SUBS to 2nd instruction of clamp code. subs will co-issue with min/max, and is further from branch. When CMP was changed to SUBS, the branch can be slower. Moving the SUBS away from the branch allows it to correctly correctly predict. PiperOrigin-RevId: 285127046
XNNPACK is a highly optimized library of floating-point neural network inference operators for ARM, WebAssembly, and x86 (SSE2 level) platforms. XNNPACK is not intended for direct use by deep learning practitioners and researchers; instead it provides low-level performance primitives for accelerating high-level machine learning frameworks, such as MediaPipe, TensorFlow Lite, and TensorFlow.js.
XNNPACK implements the following neural network operators:
All operators in XNNPACK support NHWC layout, but additionally allow custom stride along the Channel dimension. Thus, operators can consume a subset of channels in the input tensor, and produce a subset of channels in the output tensor, providing a zero-cost Channel Split and Channel Concatenation operations.
The table below presents single-threaded performance of XNNPACK library on two generations of MobileNet models and three generations of Pixel phones.
Model | Pixel, ms | Pixel 2, ms | Pixel 3a, ms |
---|---|---|---|
MobileNet v1 1.0X | 81 | 93 | 88 |
MobileNet v2 1.0X | 48 | 58 | 54 |
Benchmarked on October 9, 2019 with end2end_bench --benchmark_min_time=5
on an Android/ARM64 build (bazel build -c opt --config android_arm64 :end2end_bench
) and neural network models with randomized weights and inputs.
XNNPACK is a based on QNNPACK library. Unlike QNNPACK, XNNPACK focuses entirely on floating-point operators, and its API is no longer compatible with QNNPACK.