Abstract: This brief proposes and evaluates several OpenCL-based implementations of the Secure Hash Algorithm-3 (SHA-3) co-processor. These implementations are developed based on OpenCL optimization ...
Perform these steps to build the runtime: Clone and change to the runtime repository. Use GCC as the compiler instead of the system default (for Linux only). CC=gcc CXX=g++ cmake -G Ninja ..
Run Stable Diffusion on RK3588's Mali GPU with MLC/TVM. Currently generate a 512x512 image costs about 500 seconds (including model loading and GPU kernel compilation time. The actual inference time ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback