Tacker: Tensor-CUDA Core Kernel Fusion for Improving the GPU Utilization while Ensuring QoS.
Han ZhaoWeihao CuiQuan ChenYoutao ZhangYanchao LuChao LiJingwen LengMinyi GuoPublished in: HPCA (2022)
Keyphrases
- parallel implementation
- gpu implementation
- gpu accelerated
- parallel computing
- graphics processors
- quality of service
- graphics hardware
- real time
- parallel computation
- resource utilization
- compute unified device architecture
- general purpose
- high order
- kernel function
- resource management
- dimensionality reduction
- higher order
- graphics processing units
- parallel programming
- diffusion tensor images
- graphic processing unit
- gaussian processes
- multi sensor
- information fusion
- kernel methods
- data fusion
- response time
- optical flow