| commit | 7c8e0c7ad7643c8fa8ed0cbde0a4518c438722a6 | [log] [tgz] |
|---|---|---|
| author | Frank Barchard <fbarchard@google.com> | Sun Nov 17 00:02:36 2019 -0800 |
| committer | XNNPACK Team <xnnpack-github-robot@google.com> | Sun Nov 17 00:03:11 2019 -0800 |
| tree | 52a70c1725bf08ce479db28be9d0776f66539d17 | |
| parent | cb10f268931498907e6406ffbd8c6f74268f25ca [diff] |
4x8 IGEMM for Cortex-A53 pipelined 24.7% faster than non-pipelined kernel. Was f32_igemm_4x8__aarch64_neonfma_cortex_a75 61078821 21 Now f32_igemm_4x8__aarch64_neonfma_cortex_a53 48954181 21 PiperOrigin-RevId: 280899199
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.