Optimize applications tuned to realize the maximum performance for your SiFive RISC-V applications including neural networks, linear algebra, and signal processing
SiFive Kernel Library (SKL) is a suite of tuned routines that maximize CPU utilization, performance and power for application on SiFive RISC-V vector and matrix microarchitecture engines.
The libraries seamlessly integrate with SiFive Freedom SDK for Metal and Linux, achieving high performance workload results and include routines and kernels.
Matrix Multiplication (GEMM): 2D matrix multiplication across a variety of data types (16, 32 and 64-bit IEEE floating point, BFloat16, and quad-widening 8-bit to 32-bit integers), as well as certain 4D packed matrix layouts.
Convolution: 2D depthwise convolution kernels supporting the HWC data layout optimized for a variety of filter shapes, for convolutional neural networks (CNNs).
Nonlinear Functions: exponential function, softmax, sigmoid linear unit (SiLU), and the Gaussian error linear unit (GELU) across a similarly broad range of floating-point data types, primarily for use by or as activation functions.
Data Movement: matrix transposition and packing/unpacking functions to support the use of GEMM kernels.
Learn more about SiFive Kernel Library by downloading the product brief today or contact our sales team with any questions.
