DOT-PIM: Agile Design and Optimization Tools for Processing-In-Memory
Workshop Webpage
Please visit the DOT-PIM website https://dot-pim-workshop.com/ for more details.
About
The emerging Processing-In-Memory (PIM) architectures provide an effective solution to overcome the memory wall problem of von Neumann architectures. PIM architectures eliminate data movements between memory and computing units and are beneficial for embedded systems, and gain increasing attention from academia and industry. This workshop (DOT-PIM) will invite the leading experts from the PIM community to share their wonderful achievements in design and optimization tools for PIM. We hope DOT-PIM can help researchers to better understand the design challenges of PIM and inspire further research directions.
Program
09:00-09:05 Prof. Yu Wang, Tsinghua University, Opening
09:05-09:45 Prof. Onur Mutlu, ETH & CMU, Memory-Centric Computing
09:45-10:15 Dr. Abbas Rahimi, IBM, A Neuro-Vector-Symbolic Architecture for Data- and Compute-Efficient Continual Learning, Abstract Reasoning, and Combinatorial Inference
10:15-10:30 Coffee Break
10:30-11:00 Prof. Huaqiang Wu, Tsinghua University, Full-stack Deployment and Design tools for RRAM-based Compute-in-memory System
11:00-11:30 Prof. Li Jiang, Shanghai Jiao Tong University & Huawei, Opportunities and Challenges for Process-In-Memory (PIM) Technology in ICT Products
11:30-12:00 Prof. Xiaoming Chen, ICT, Chinese Academy of Sciences, EDA Toolchain for Processing-in-Memory CNN Accelerators
12:00-13:30 Lunch & Poster Sharing Sessions
13:30-14:10 Prof. Yu (Kevin) Cao, University of Minnesota, HISIM: Heterogeneous Integration Simulator with 2.5D/3D Interconnect Modeling
14:10-14:40 Prof. Fengbin Tu, HKUST, Digital Computing-In-Memory Architecture and Design Automation
14:40-15:10 Prof. Grace Li Zhang, TU Darmstadt, Robust and Efficient Analog In-Memory-Computing Platforms for Neural Networks
15:10-15:25 Coffee Break
15:25-15:55 Prof. Yaoyu Tao, Peking University, Fast and Reconfigurable Sort-In-Memory System Enabled by Memristors
15:55-16:20 Mr. Yuan-Chun Luo, GIT, advisor: Prof. Shimeng Yu, Benchmarking Framework for Non-volatile Capacitive Compute-in-Memory
16:20-16:45 Mr. Shiyu Li, Duke, advisor: Prof. Yiran Chen, Towards Efficient Processing in Memory AI Systems with Cross-layer Optimization
16:45-16:50 Closing Remark
Organizers
Wang Yu
Electronic Engineering Department, Tsinghua University, Beijing, China
