iGniter: Interference-Aware GPU Resource Provisioning for Predictable DNN Inference in the Cloud.
Fei XuJianian XuJiabin ChenLi ChenRuitao ShangZhi ZhouFangming LiuPublished in: CoRR (2022)
Keyphrases
- cloud computing
- real time
- probabilistic inference
- inference process
- bayesian inference
- parallel computing
- graphics hardware
- cluster of workstations
- bayesian networks
- gpu implementation
- inference mechanism
- belief networks
- bayesian model
- training process
- graphics processors
- parallel programming
- computing platform
- multipath
- parallel processing
- probabilistic model
- database systems
- computer vision