This repository provides the public references for accelerating AI SOC with high energy-efficient performance. The main applications are focusing on HPC (high performance computing), NAS (neural architecture search) and low-power edge computing.
- 2018 | BISMO: A Scalable Bit-Serial Matrix Multiplication Overlay for Reconfigurable Computing | NTNU Group | FPL |
PDF
- 2018 | Bit Fusion: Bit-Level Dynamically Composable Architecture for Accelerating Deep Neural Networks | H. Sharma, et al. | ISCA |
PDF
- 2018 | Harnessing GPU Tensor Cores for Fast FP16 Arithmetic to Speed up Mixed-Precision Iterative Refinement Solvers | A. Haidar, et al. | Supercomputing |
PDF
- 2018 | Mixed-precision in-memory computing | Manuel Le Gallo, et al. | Nature Electronics |
PDF
- 2018 | DNPU: An Energy-Efficient Deep-Learning Processor with Heterogeneous MultiCore Architecture | KAIST Group | Micro |
PDF
look-up Table
- 2018 | DNPU: An 8.1TOPS/W Reconfigurable CNN-RNN Processor for General-Purpose Deep Neural Networks | KAIST Group | ISSCC |
PDF
look-up Table
- 2018 | Efficient Fixed/Floating-Point Merged Mixed-Precision Multiply-Accumulate Unit for Deep Learning Processors | H. Zhang, et al. | ISCAS |
PDF
- 2018 | Mixed Precision Training | Baidu | ICLR |
PDF
- 2017 | Bit-Pragmatic Deep Neural Network Computing | UToronto Group | Micro |
PDF
- 2016 | Stripes: Bit-Serial Deep Neural Network Computing | UToronto Group | Micro |
PDF
-
2015 | Performance Analysis of Karatsuba Multiplication Algorithm for Different Bit Lengths | C. Eyupoglu |
PDF
-
2015 | An Efficient Floating Point Multiplier Design for High Speed Applications using Karatsuba Algorithm and Urdhva-Tiryagbhyam Algorithm | S. Arish, et al. | ICSC |
PDF
-
2014 | An Efficient Baugh-Wooley Multiplication Algorithm for 32-bit Synchronous Multiplication |
PDF
-
2013 | Implementation of Baugh-Wooley Multiplier Based on Soft-Core Processor |
PDF
-
2012 | An Efficient Baugh-Wooley Architecture for Both Signed & Unsigned Multiplication |
PDF
-
2012 | A High Speed Wallace Tree Multiplier Using Modified Booth Algorithm for Fast Arithmetic Circuits |
PDF
-
201x | title | author, et al. | conf. or Journal |
PDF
- 2019 | HAQ: Hardware-Aware Automated Quantization with Mixed Precision | Song Han Group | CVPR |
PDF