The CRNCH 2024 Summit was be held on February 8th-9th. Please check the official site for more detail. This site contains links to slides for those presenters who provided them.
Thursday, February 8th, 2024
Time (EST) | Session | Speaker/Talk Title |
---|---|---|
8:30 AM | Registration and Continental Breakfast | |
9:00 AM | Keynote | Hal Finkel, DoE, director of ASCR [Bio] |
Session chair – Hyesoon Kim | The Department of Energy’s Advanced Scientific Computing Research Program: Recent Advances and Trends Motivating Future Research | |
10:00 AM | BREAK | |
10:20 AM | Session 1: ML Systems | Alexey Tumanov, Georgia Tech [Bio] |
Session chair – Tushar Krishna | Serving Pareto Frontiers for Goodput-Sensitive Applications [Abstract] | |
Larry Heck, Georgia Tech [Bio] [Abstract] | ||
A Brief History of Large Language Models… and a glimpse into the future | ||
Michael Isaev, Georgia Tech [Bio] | ||
Calculon: a methodology and tool for high-level co-design of systems and large language models [Abstract] | ||
Pan Li, Georgia Tech [Bio] | ||
Machine learning on Graphs — system-friendly algorithm design [Abstract] | ||
12:00 PM | Lunch / “Food for Thought” | Santosh Vempala, GT Computer Science [Bio] |
Session chair – Rich Vuduc | Emergent Computation and Learning in a Model of the Brain: from Neurons and Synapses to Memory and Cognition [Abstract] | |
1:20 PM | Plenary Talk | Prasanna Balaprakash, ORNL Director of AI Programs [Bio] |
Session chair – Rich Vuduc | Taming High-Performance Computing Platform Complexity with Machine Learning [Abstract] | |
2:20 PM | BREAK | |
2:35 PM | Session 2: Extreme Low Power Computing | Josiah Hester, GT CS/IC [Bio] |
Session Chair – Celine Lin | Embracing Failure: Axioms and Future Directions of Intermittent Computing [Abstract] | |
Azad Naeemi, GT ECE [Bio] | ||
Novel Ternary Content Addressable Memory Designs Based on Emerging Magnetic and Magnetoelectric Devices [Abstract] | ||
3:50 PM | Session 3: Simulation and Testbeds | Jason Lowe-Power, UC Davis [Bio] |
Session Chair – Hyesoon Kim | Leveraging the open-source ethos: How and why gem5 is being used to design the next generation of supercomputers [Abstract] | |
Jeff Young, Georgia Tech [Bio] | ||
Novel Architecture Simulation, Emulation, and Evaluation with the CRNCH Rogues Gallery | ||
Murali Emani, Argonne National Lab [Bio] | ||
Accelerating Scientific Machine Learning with AI Accelerators at ALCF AI Testbed | ||
4:40 PM | Poster Session | TSRB event space – See poster abstracts and pdfs at this link |
Session Chair – Jeff Young | ||
6:00 PM | End of Day 1 – Time for Networking and Discussions |
Friday, February 9rd, 2024
Time (EST) | Session | Speaker/Talk Title |
---|---|---|
9:00 AM | Keynote | Yiran Chen, Duke University [Bio] |
Session Chair – Celine Lin | Optimizing AI Systems through Cross-Layer Design: A Data-Centric Approach [Abstract] | |
10:00 AM | Session 4: Analog and Neuromorphic | Jennifer Hasler, Georgia Tech ECE [Bio] |
Session Chair – Jeff Young | A Vision for Neuromorphic Hardware Development Tools | |
Benjamin Feinberg, Sandia [Bio] | ||
Where Is My Analog Accelerator? [Abstract] | ||
10:50 AM | BREAK | |
11:10 AM | Session 5: Emerging Applications | Spencer Bryngelson, GT CSE [Bio] |
Session Chair – Rich Vuduc | Hybrid algorithm for incompressible flow simulation on quantum hardware [Abstract] | |
Mathias Jacquelin, Cerebras [Bio] | ||
Multi-Dimensional Seismic Processing with Algebraic Compression on Cerebras CS-2 Systems | ||
12:00 PM | Lunch Panel | H/W and S/W co-design panel Vivek Sarkar (GT), Roberto Gioiosa (PNNL), Yiran Chen (Duke), Mathias Jacquelin (Cerebras), Hyesoon Kim (GT), Peter Kogge (UND) |
Moderator – Tushar Krishna | ||
1:20 PM | Session 6: HW-SW Co-Design for ML | Roberto Gioiosa, PNNL [Bio] |
Session Chair – Hyesoon Kim | Hardware Accelerators: Use, Misuse, and Co-Design | |
Bahar Asgari, UMD [Bio] | ||
Exploiting Diversity in Sparse Problems to Enable Efficient Computing | ||
2:30 PM | Closing Remarks |