Open House and Demonstration I

Demonstration

1. Application-specific Compression Hardware Co-design

     Developers: Xuejiao LIU, Yu LIU, Xianghong HU, Haowei ZHANG, Xijie HUANG, Man To YUNG
     Presenter: Man To YUNG

2. Co-design Toolchain: Quantization, Compilation, and Simulation

     Developers: Haowei Zhang, Yu LIU
     Presenter: Haowei ZHANG

3. SRAM-based Compute-In-Memory

     Developers: Xiaomeng WANG, Xuejiao LIU, Xihao GUAN, Patrick KONG
     Presenter: Xihao GUAN

Posters by ACCESS Research Team

4. A Scalable Bit-level Accelerator with Supporting Pre-layer Mixed-precision Quantization

     Presenter: Jianwei ZHENG

5. WASP: Efficient Power Management Enabling Workload-Aware, Self-Powered AIoT Devices

     Presenter: Xiaofeng HUO

6. Training Integer Neural Networks Using Integer-only Arithmetic

     Presenter: Maolin WANG

Posters by ACCESS Students

7. Random Memristor-based Convolutional Echo State Network for Human Activity Recognition

     Member(s): Shaocong WANG, Hegan CHEN (HKU)
     Advisor(s): Zhongrui WANG (HKU)

8. Adaptive Quantization for In-memory Computing with SOT-MRAM

     Member(s): Zhihua XIAO (HKUST)
     Advisor(s): Qiming SHAO (HKUST)

9. Step RRAM – MLC RRAM with Fast Write/Verify Convergence and Its Energy, Area, and Access Time Analysis

     Member(s): Wei-Chen CHEN, Luke UPTON (Stanford)
     Advisor(s): Philip WONG, Boris MURMANN (Stanford)

10. PHANES: ReRAM-based Photonic Accelerators for Deep Neural Networks

        Member(s): Jiaqi LIU, Chengao SHI, Chengeng LI, Lin CHEN (HKUST)
        Advisor(s): Jiang XU (HKUST)

11. NARX Neural Network-assisted ReRAM Modeling

        Member(s): Zhao RONG (HKUST)
        Advisor(s): Mansun CHAN (HKUST)

12. Compression of Generative Pre-trained Language Models via Quantization

        Member(s): Chaofan TAO (HKU)
        Advisor(s): Ngai WONG, Ping LUO (HKU)

13. Improving the Energy Efficiency and Robustness of tinyML Computer Vision using Log-Gradient Input Images

        Member(s): Qianyun LU (Stanford)
        Advisor(s): Boris MURMANN (Stanford)

14. CHIMERA: Efficient DNN Inference and Training at the Edge with On-Chip Resistive RAM

        Member(s): Kartik PRABHU, Massimo GIORDANO, Kalhan KOUL, Robert RADWAY, Albert GURAL, Rohan DOSHI, Zainab KHAN,
                             John KUSTIN, Timothy LIU, Gregorio LOPES, Victor TURBINER (Stanford)
        Advisor(s): Priyanka RAINA, Boris MURMANN, Subhasish MITRA (Stanford)

Remarks: All demonstrations and posters will be available on-site both Day 1 (June 9) and Day 2 (June 10).