MLBench’25 (https://memani1.github.io/mlbench25/): Fifth Workshop on Benchmarking and Performance of Machine Learning Workloads on Emerging Hardware
In conjunction with ACM ASPLOS and co-located with EuroSys
March 30, 2025 • Rotterdam, Netherlands

Overview
The Fifth Workshop on Benchmarking and Performance of Machine Learning Workloads on Emerging Hardware (MLBench’25) is a premier forum for presenting and discussing research results at the intersection of machine learning workloads, emerging hardware platforms, and performance evaluation methodologies. MLBench’25 aims to bring together experts from academia, industry, and national labs to share novel insights into designing, benchmarking, and tuning machine learning systems at scale.

Key problems that we seek to address are:
(i) which representative ML benchmarks cater to workloads seen in industry, national labs, and interdisciplinary sciences;
(ii) how to characterize the ML workloads based on their interaction with hardware;
(iii) which novel aspects of hardware, such as heterogeneity in compute, memory, and networking, will drive their adoption;
(iv) performance modeling and projections to next-generation hardware.

MLBench’25 encourages both full papers (8–10 pages) and short/position papers (4–6 pages) describing new ideas, experimental studies, theoretical results, and system prototypes related to benchmarking and performance. Alongside accepted papers, the workshop program will feature invited talks and expert panels highlighting cutting-edge research and industry insights. MLCommons will sponsor a Best Student Paper Award for an outstanding submission.

Important Dates (AoE)
Submission Deadline: February 11, 2025
Acceptance Notification: February 21, 2025
Camera-Ready Deadline: March 14, 2025
Workshop Date: March 30, 2025
All deadlines are Anywhere on Earth (AoE) and are firm.