Burnt Sienna Team 2022
Xuesi Chen, Daniel Ernst, Margret Riegert
Hotspot Mitigation for Systolic Array CNN Accelerators
On-chip thermal hotspots have become an increasing concern in recent generations of ML accelerators. With the increasing demand for neural network processing from ML applications such as natural language processing or image recognition, and deeper and larger emerging neural networks, a software tool that can accurately predict the appearance of thermal hot-spots has the potential to improve thermal performance of future ML accelerators. Our project, NNShim, builds upon the existing systolic ML simulator SCALE-Sim for data movement and the thermal hotspot simulator HotGauge to generate temperature and power maps, with a specific neural network model and hardware configuration. Our simulation results prove that decreasing clock frequency, increasing aspect ratios and applying row and column skipping are effective ways to cool ML accelerators and mitigating hot-spots.
Related Tech Notes
- Investigation on the Power Consumption Models for ML Accelerator by Xuesi Chen
- Sparsity in Neural Networks by Daniel Ernst
Project Poster
- Senior Capstone Projects Summary for the 2021-22 Academic Year > Burnt Sienna Team 2022
Search the Handbook:
Handbook Overview
- Introduction and Acknowledgements
- Senior Capstone Projects Summary for the 2022-23 Academic Year
- Senior Capstone Projects Summary for the 2021-22 Academic Year
- Senior Capstone Projects Summary for the 2020-21 Academic Year
- Senior Capstone Projects Summary for the 2019-20 Academic Year
- Senior Capstone Projects Summary for the 2018-19 Academic Year
- Senior Capstone Projects Summary for the 2017-18 Academic Year
- Senior Capstone Projects Summary for the 2016-17 Academic Year
- Senior Capstone Projects Summary for the 2015-16 Academic Year
- Senior Capstone Projects Summary for the 2014-15 Academic Year
- Senior Capstone Projects Summary for the 2013-14 Academic Year
- Senior Capstone Projects Summary for the 2012-13 Academic Year
- 1. Design Process
- 2. Management
- 3. Technologies
- 4. Communications And Life Skills
- 5. Tech Notes
Top Topics
Apple iPhone Assistive Technologies Big Data Bridge Inspection Business Strategy Communications Computer Vision Consumer Technologies Cough detection Creativity & Innovation Data Analysis Design for X Emergency resource management Emerging Technologies Encryption Engineering Economics Ethics GPS Indoor Navigation Industrial Technologies Internet of Things Interpersonal Skills Legal & Intellectual Property Marketing & Customer Research Mobile Applications OpenCV Product Development Life Cycle Product Liability Prototyping & Manufacturing Recommender System Remote Keyless Entry Risk Risk Management Security Sensors Signal Processing Societal Impact Synthetic Aperture Radar Tuberculosis UAV Drones UAV Motion Unmanned Technologies Virtual Reality Visual Impairment Wireless