| A | B | C | D | E | F | G | |
|---|---|---|---|---|---|---|---|
1 | Section | Time (EST) | Title | Speaker | Attendance | Workshop | Session Chair |
2 | Opening | 9:00AM - 9:05 AM | Opening and Announcements | Amir Yazdanbakhsh (Google) | |||
3 | Keynote Speaker | 9:05AM - 9:50AM | ML4ML: The Intriguing Interplay between Moore’s Law & Machine Learning | Parthasarathy Ranganathan (Google) | Neeraja J. Yadwadkar (UT Austin) | ||
4 | 9:50AM - 10:00AM | Break | |||||
5 | Session 1 | 10:00AM - 10:10AM | Lightweight ML-based Runtime Prefetcher Selection on Many-core Platforms | Erika S Alcorta | In-Person | MLArchSys | Amir Yazdanbakhsh (Google DeepMind) |
6 | 10:10AM - 10:20AM | DOSA: One-Loop DSE for DNN Accelerators Using Differentiable Models | Charles Hong | In-Person | MLArchSys | ||
7 | 10:20AM - 10:30AM | Towards Efficient Multi-Agent Learning Systems | Kailash Gogineni | In-Person | MLArchSys | ||
8 | 10:30AM -10:40AM | ParaGAN: A Cloud Training Framework for Generative Adversarial Networks | Ziji Shi | In-Person | MLArchSys | ||
9 | 10:40AM - 10:50AM | Sample-Efficient Mapspace Optimization for DNN Accelerators with Bayesian Learning | Grace Dinh | In-Person | MLArchSys | ||
10 | 10:50AM - 11:00AM | Online Learning for Right-Sizing Serverless Functions | Prasoon Sinha | In-Person | MLArchSys | ||
11 | 11:00AM - 11:05AM | Buffer | |||||
12 | 11:05PM - 11:20PM | Coffee Break | |||||
13 | Keynote Speaker | 11:20AM - 12:05PM | LLM Training at Wafer-Scale | Valavan Manohararajah (Cerebras) | Bahar Asgari (UMD) | ||
14 | Session 2 | 12:05PM - 12:15PM | Scaling Infrastructure to Support Multi-Trillion Parameter LLM Training | Mikhail Isaev | In-Person | ASSYST | |
15 | 12:15PM - 12:25PM | Accuracy Boosters: Epoch-Driven Mixed-Mantissa Block Floating Point for DNN Training | Simla Burcu Harma | In-Person | MLArchSys | ||
16 | 12:25PM - 12:30PM | Buffer | Buffer | ||||
17 | 12:30PM - 1:40PM | Lunch | |||||
18 | Keynote Speaker | 1:40PM - 2:25PM | Faster Neural Network Training, Algorithmically | Jonathan Frankle (MosaicML) | Tushar Krishna (GaTech) | ||
19 | Session 3 | 2:25PM - 2:35PM | Accelerating Attention Based Models via HW-SW Co-Design using Fine-Grained Sparsification | Abhimanyu Rajesh Bambhaniya | In-Person | ASSYST | |
20 | 2:35PM - 2:45PM | Efficient Deployment of Transformer Models on Edge TPU Accelerators: A Real System Evaluation | Mohammadreza Mohammadi | In-Person | ASSYST | ||
21 | 2:45PM - 2:55PM | TAP: Efficient Derivation of Tensor Parallel Plans for Large Neural Networks | Ziji Shi | In-Person | ASSYST | ||
22 | 2:55PM - 3:05PM | Full Stack Optimization of Transformer Inference | Coleman Hooper | In-Person | ASSYST | ||
23 | 3:05PM - 3:15PM | Towards A Reconfigurable Systolic Array with Multi-Level Packing for Transformers | Tiandong Zhao | In-Person | ASSYST | ||
24 | 3:15PM - 3:25PM | A Metric Driven Approach to Mixed Precision Training | Mitchelle Rasquinha | In-Person | ASSYST | ||
25 | 3:25PM - 3:30PM | Buffer | |||||
26 | 3:30PM - 4:00PM | Coffee Break | |||||
27 | Keynote Speaker | 4:00PM - 4:45PM | Software, Hardware, and Model Codesign for High-performance Transformer-based Large Models | Zongwei Zhou (Google) | Suvinay Subramanian (Google) | ||
28 | Closing Remarks | 4:45PM - 4:50PM | Closing Remarks | Suvinay Subramanian (Google) | |||
29 | Poster Session | 4:50PM - 5:30 PM | Poster Session | ||||