commit | 9cdade312a0b9df393dfaeab377e81334aa04658 | [log] [tgz] |
---|---|---|
author | Erich Elsen <eriche@google.com> | Wed Oct 16 05:26:59 2019 -0700 |
committer | XNNPACK Team <xnnpack-github-robot@google.com> | Wed Oct 16 05:27:19 2019 -0700 |
tree | 83edcbe0c71f65ea4e82e670fd6cfad31c88f5f9 | |
parent | 34dc2c0784c4efea9b3c70b7f0a68933e2c5a009 [diff] |
Add prefetch instructions to 16x1, 16x2, 16x4 kernels. Switch to 16x1 pipelined kernel from vanilla 16x1. PiperOrigin-RevId: 275010952
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 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. However, unlike QNNPACK, XNNPACK focuses entirely on floating-point operators, and its API is no longer compatible with QNNPACK.