Romou: rapidly generate high-performance tensor kernels for mobile GPUs.
Rendong LiangTing CaoJicheng WenManni WangYang WangJianhua ZouYunxin LiuPublished in: MobiCom (2022)
Keyphrases
- graphics processing units
- mobile phone
- kernel function
- automatically generate
- support vector
- mobile devices
- general purpose
- higher order
- highly parallel
- machine learning
- high order
- mobile computing
- computational power
- tensor decomposition
- parallel programming
- diffusion tensor
- mobile networks
- mobile applications
- linear combination
- dimensionality reduction