iGniter: Interference-Aware GPU Resource Provisioning for Predictable DNN Inference in the Cloud.
Fei XuJianian XuJiabin ChenLi ChenRuitao ShangZhi ZhouFangming LiuPublished in: IEEE Trans. Parallel Distributed Syst. (2023)
Keyphrases
- cloud computing
- real time
- bayesian networks
- bayesian inference
- inference mechanism
- gpu accelerated
- probabilistic reasoning
- training process
- probabilistic inference
- general purpose
- computer vision
- data sets
- conditional random fields
- parallel processing
- belief networks
- computer simulation
- multipath
- parallel implementation
- parallel computing
- neural network
- graphics processors
- database