DELTA: Dynamically Optimizing GPU Memory beyond Tensor Recomputation.
Yu TangChenyu WangYufan ZhangYuliang LiuXingcheng ZhangLinbo QiaoZhiquan LaiDongsheng LiPublished in: CoRR (2022)
Keyphrases
- real time
- higher order
- high order
- main memory
- memory usage
- intel xeon
- data transfer
- parallel processing
- memory requirements
- random access
- multithreading
- graphics processors
- computational power
- dynamically created
- parallel computing
- dimensionality reduction
- graphics hardware
- multi threaded
- message passing
- computing power
- memory space
- memory size
- data structure