A 95.6-TOPS/W Deep Learning Inference Accelerator With Per-Vector Scaled 4-bit Quantization in 5 nm.
Ben KellerRangharajan VenkatesanSteve DaiStephen G. TellBrian ZimmerCharbel SakrWilliam J. DallyC. Thomas GrayBrucek KhailanyPublished in: IEEE J. Solid State Circuits (2023)
Keyphrases
- deep learning
- successive approximation
- unsupervised learning
- machine learning
- uniform quantization
- unsupervised feature learning
- weakly supervised
- bayesian networks
- deep architectures
- mental models
- feature vectors
- information retrieval
- image processing
- domain specific
- decision support system
- restricted boltzmann machine