Lab Seminar 2019

Information

• Time: 7pm Wednesdays
• Place: 106-dong 511

Schedule

Day	Date	Presenter	Contents	Note
1	Tue 6/18, 5pm	Sugil	DAC papers: Noise injection adaptation, ApproxLP, E-LSTM
		Hyeonuk	DAC papers: 41_2, SkippyNN (SC), BitBlade, ABM
2	Wed 6/26	Jongeun	DAC Tutorial 5
3	Fri 7/12, 1pm	Jun-gi	SOSPCS

Paper list

1. https://arxiv.org/pdf/1811.07755.pdf (UGConv)
2. dynamic pruning
   • https://papers.nips.cc/paper/6813-runtime-neural-pruning
   • https://arxiv.org/abs/1810.05331
   • https://arxiv.org/abs/1805.12549
   • attention mechanism paper(s)
3. DAC
   • 41_2.pdf Acceleration of DNN Backward Propagation by Selective Computation of Gradients
   • 41_1.pdf BitBlade: Area and Energy-efficient Precision-scalable Neural Network Accelerator with Bitwise Summation
   • 41_4.pdf ABM-spConv: A Novel Approach to FPGA-based Acceleration of Convolutional Neural Network Inference
   • Noise Injection Adaption: End-to-End ReRAM Crossbar Non-ideal Effect Adaption for Neural Network Mapping
   • ApproxLP: Approximate Multiplication with Linearization and Iterative Error Control
     • How to extend it to other ranges or other types (eg, fixed-point)?
   • E-LSTM: Efficient Inference of Sparse LSTM on Embedded Heterogeneous System
   • SkippyNN: An Embedded Stochastic-computing Accelerator for Convolutional Neural Networks.
4. SOSPCS: (TVLSI 2019 June)
   • "Self-Optimizing and Self-Programming Computing Systems: A Combined Compiler, Complex Networks, and Machine Learning Approach", Xiao et al.