Summary of Safepath: a System-level Approach For Efficient Power and Thermal Estimation Of Convolutional Neural Network Accelerator, by Yukai Chen et al.
SAfEPaTh: A System-Level Approach for Efficient Power and Thermal Estimation of Convolutional Neural Network Accelerator
by Yukai Chen, Simei Yang, Debjyoti Bhattacharjee, Francky Catthoor, Arindam Mallik
First submitted to arxiv on: 24 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Performance (cs.PF)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary SAfEPaTh is a novel system-level approach that estimates power and temperature in tile-based Convolutional Neural Network (CNN) accelerators with high accuracy. It addresses both steady-state and transient-state scenarios by capturing the dynamic effects of pipeline bubbles in interlayer pipelines using real CNN workloads. This methodology eliminates the need for circuit-level simulations or on-chip measurements, leveraging TANIA, a hybrid digital-analog tile-based accelerator featuring analog-in-memory computing cores alongside digital cores. Simulation results using the ResNet18 model demonstrate SAfEPaTh’s capability to accurately estimate power and temperature within 500 seconds, encompassing CNN model accelerator mapping exploration and detailed power and thermal estimations. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary SAfEPaTh is a new way to measure how much energy and heat a special kind of computer chip uses when it runs certain types of AI programs. It can predict how the chip will use energy and produce heat in different situations, which helps designers make sure the chip doesn’t get too hot or use too much power. This makes it easier to design chips that work well while also being efficient and safe. |
Keywords
» Artificial intelligence » Cnn » Neural network » Temperature
