From ddfac7b54efa808d4acb4e9d066ceb52a713e311 Mon Sep 17 00:00:00 2001 From: Liangliang He Date: Tue, 22 May 2018 19:04:42 +0800 Subject: [PATCH] Update README.md --- README.md | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index b71af624..520b5de0 100644 --- a/README.md +++ b/README.md @@ -12,12 +12,17 @@ mobile heterogeneous computing platforms. The design is focused on the following targets: * Performance - * The runtime is highly optimized with NEON, OpenCL and Hexagon. Except for the - inference speed, the initialization speed is also intensively optimized. + * The runtime is highly optimized with NEON, OpenCL and Hexagon, and + [Winograd algorithm](https://arxiv.org/abs/1509.09308) is introduced to + speed up the convolution operations. Except for the inference speed, the + initialization speed is also intensively optimized. * Power consumption - * Chip dependent power options are included as advanced API. + * Chip dependent power options like big.LITTLE scheduling, Adreno GPU hints are + included as advanced API. * Memory usage and library footprint * Graph level memory allocation optimization and buffer reuse is supported. + The core library tries to keep minium external dependencies to keep the + library footprint small. * Model protection * Model protection is one the highest priority feature from the beginning of the design. Various techniques are introduced like coverting models to C++ -- GitLab