| commit | 263bb090aa5f04110972e34dde82f0f46639eec9 | [log] [tgz] |
|---|---|---|
| author | Frank Barchard <fbarchard@google.com> | Mon Oct 28 15:28:46 2019 -0700 |
| committer | XNNPACK Team <xnnpack-github-robot@google.com> | Mon Oct 28 15:40:18 2019 -0700 |
| tree | fe018723b045cd077758122b91efa330ad108289 | |
| parent | feb49230bef35ecaf1a74d3e6c732bd5151ec433 [diff] |
Cortex A76 use 6x8 micro kernel On big core A75 kernel is 0.3% faster with 6x8 than 4x8 A53 kernel is 11.2% faster with 6x8 than 4x8 On little core A75 kernel is 6.4% faster with 6x8 than 4x8 A53 kernel is 9.2% faster with 6x8 than 4x8 PiperOrigin-RevId: 277156128
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.