As Machine Learning (ML) continues to permeate all areas of computing, software system designers and developers are increasingly adopting ML-based solutions to tackle complex challenges, particularly in optimization and hardware design. ML is being leveraged to address a wide array of problems, including the design of cost models, code optimization heuristics, efficient search space exploration, automatic optimization, and program synthesis. The development of accurate ML models, feature engineering, verification and validation of results, and the selection and curation of representative training data are all significant, ongoing challenges in this field. These topics are actively explored by a large community of researchers in both industry and academia. MLS/H offers an excellent venue for the international research community to exchange ideas and techniques, focusing on the application of machine learning to system challenges, especially within the software stack and hardware domains.

For more information, please visit our website.

Submission Guidelines

We invite speakers from a variety of institutions, including academia, research institutes, and industry. Please submit your abstract, optionally along with your final presentation slides if available. We are seeking speakers for technical talks (40 minutes in length). The submitted abstracts will be reviewed by the organizers and program committee members, and the slides for accepted talks will be published in our online proceedings. Please submit your abstract using this link.

List of Topics

We invite speakers interested in presenting their work on topics including, but not limited to, the following areas:

ML for the Software Stack

• Heuristics and cost model construction
• Optimization space exploration
• Automatic code optimization
• Bug detection
• Program synthesis
• Program and code representation
• Significant training paradigms

ML for Hardware

• ML models for optimal FPGA configuration
• Load balancing between CPUs and accelerators (e.g., GPUs, TPUs)
• ML models to improve computer architecture design
• Analysis and techniques for defining meaningful representations (features) for compilers and hardware

Training Data

• Exploring the availability or generation of efficient training data for compilers and hardware
• Utilizing graph-based data for machine learning
• Improving training data quality

We also welcome submissions related to ML compilers, on-device training, and the use of specialized ML hardware!

Committees

Program Committee

• TBD

Organizing committee

• Eun Jung (EJ) Park (Qualcomm Inc).
• Riyadh Baghdadi (New York University Abu Dhabi and Massachusetts Institute of Technology).
• Joseph Manzano (Pacific Northwest National Laboratory).

Venue

Long Beach, CA, USA

Contact

eunjpark(at)qualcomm.com