The widespread adoption of mobile devices has revolutionized the way people interact with technology. Mobile platforms are becoming increasingly capable of running sophisticated artificial intelligence (AI) applications, including generative AI models. Generative AI, such as large language models and stable diffusion, has enabled remarkable advancements in various domains, including computer vision, natural language processing, and music generation. However, deploying and running these resource-intensive generative AI models efficiently on mobile platforms pose significant challenges.
This workshop aims to bring together researchers, practitioners, and industry experts to explore the architecture and system support required for effectively deploying generative AI models on mobile platforms. We will discuss the unique constraints of mobile devices, including limited computational resources, energy constraints, and storage limitations. Additionally, we will address the challenges of training and fine-tuning generative AI models on mobile devices while ensuring privacy and security.
56th IEEE/ACM International Symposium on Microarchitecture (MICRO 2023)
We prepare invited talks and also invite submissions on, but not limited to, the following topics:
Mobile-specific architectural designs for generative AI models
System-level optimizations for efficient execution of generative AI on mobile devices
Energy-efficient techniques for training and inference of generative AI models
Model compression methods, such as sparsification and quantization, for reducing the memory footprint of generative AI on mobile platforms
Federated learning and collaborative techniques for training generative AI models on distributed mobile devices
Privacy-preserving and secure frameworks for generative AI on mobile platforms
Integration of generative AI into mobile applications and services
Case studies and real-world deployments of generative AI on mobile platforms
Wild and crazy ideas (WACI) of generative AI use cases
We invite researchers and practitioners to submit original papers on their ongoing research related to architecture and system support for generative AI on mobile platforms. Submissions should follow the MICRO 2023 conference format and must not exceed 4 pages, including references. All submissions will undergo a peer-review process by the workshop's program committee.
Submission date: 9/4 AoE (Monday)
Notification date: Tentative 9/22 (Friday)
A submission site on OpenReview: https://openreview.net/group?id=ACM.org/MICRO/2023/Workshop/SAGE
If you have questions, please feel free to contact us via micro.sage2023@gmail.com
The workshop will consist of the following sessions:
Keynote presentations by renowned experts in the field of generative AI and mobile computing.
Academic and industry panels to discuss the challenges, advancements, and future directions in architecture and system support for generative AI on mobile platforms.
Paper presentations and discussions on ongoing research, recent breakthroughs, and novel approaches related to the workshop's theme.
Interactive sessions and demos to showcase mobile-based generative AI systems, frameworks, and applications.
Hongil Yoon, Google
Amir Yazdanbakhsh, Google DeepMind
Yingyan (Celine) Lin, Georgia Institute of Technology
Vijay Janapa Reddi, Harvard University
Jae W. Lee, Seoul National University