MonoNN: Enabling a New Monolithic Optimization Space for Neural Network Inference Tasks on Modern GPU-Centric Architectures.
Donglin ZhuangZhen ZhengHaojun XiaXiafei QiuJunjie BaiWei LinShuaiwen Leon SongPublished in: OSDI (2024)
Keyphrases
- neural network
- image reconstruction from projections
- real time
- global optimization
- optimization algorithm
- optimization method
- low dimensional
- artificial neural networks
- parallel computing
- optimization process
- self organizing maps
- fault diagnosis
- general purpose
- optimization problems
- neural network model
- probabilistic inference
- search space
- constrained optimization
- high dimensional
- highly non linear
- neural network is trained
- heterogeneous computing
- genetic algorithm