[MLBench'26] The Sixth Workshop on
Benchmarking and Performance of Machine Learning Workloads on Emerging Hardware

To be held with ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) .
March 22, 2026


Pittsburgh, PA, USA

About

With evolving system architectures, hardware and software stacks, diverse machine learning (ML) workloads, and data, it is important to understand how these components interact with each other. Well-defined benchmarking procedures help evaluate and reason the performance gains with ML workload-to-system mappings. In this MLBench workshop, we welcome all novel submissions in benchmarking machine learning workloads from all disciplines, such as image and speech recognition, language processing, drug discovery, simulations, and scientific applications.

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.

Along with selected publications, the workshop program will also have experts in these research areas presenting their recent work and potential directions to pursue.

The program details from the previous workshops held are at ASPLOS'25 MLSys'23, MLSys'22, MLSys'21 and MLSys'20.

Schedule

8:00 AM - 8:45 AM: Toward Sustainable Data Centers for Artificial Intelligence
Benjamin Lee (University of Pennsylvania)

8:45 AM - 9:10 AM: CPU-Centric Agentic AI Characterization
Ritik Raj, Sarbartha Banerjee (Georgia Institute of Technology), Hong Wang (Intel), Tushar Krishna (Georgia Institute of Technology)

9:10 AM - 9:35 AM: Retrieval Reliability using Hardware Profiling in Long-Context LLM Inference
Srinija Ramichetty (George Washington University), Mahmoud Abumandour, Alaa Alameldeen (Simon Fraser University), Guru Venkataramani (George Washington University)

9:35 AM - 10:00 AM: CTS-Bench: Benchmarking Graph Coarsening Trade-offs for GNNs in Clock Tree Synthesis
Barsat Khadka (University of Southern Mississippi), Roxy Kawsher Ahmed (Intel), Md Rubel Ahmed (Louisiana Tech University)

10:00 AM - 10:30 AM: Coffee Break

10:30 AM - 11:15 AM: Characterizing and Enhancing the Resilience of Generative Large Language Models against Soft Errors
Lishan Yang (George Mason University)

11:15 AM - 11:40 AM: End-to-End Modeling and Optimization of Multi-Stage LLM Serving
Abhimanyu Bambhaniya, Hanjiang Wu (Georgia Institute of Technology), Suvinay Subramanian (Google), Sudarshan Srinivasan (Intel), Amir Yazdanbakhsh (Google DeepMind), Midhilesh Elavazhagan, Madhu Kumar (Intel), Tushar Krishna (Georgia Institute of Technology)

11:40 AM - 12:00 PM: Analyzing Latency Hiding and Parallelism in an MLIR-Based AI Kernel Compiler
Mohammed Javed Absar, Samarth Narang, Iulian V. Brumar, Mitesh Kothari, Zachary Zipper, Muthu Baskaran (Qualcomm)

12:00 PM - 1:30 PM: Lunch

1:30 PM - 2:15 PM: KForge - Program Synthesis for Diverse AI Hardware Accelerators
Ankita Nayak (Gimlet Labs)

2:15 PM - 3:00 PM: Low Overhead Continuous Profiling of AI Services
Keren Zhou (George Mason University, OpenAI)

3:00 PM - 3:25 PM: SCALE-Sim TPU: Validating and Extending SCALE-Sim for TPUs
Jingtian Dang, Ritik Raj, Changhai Man, Jianming Tong, Tushar Krishna (Georgia Institute of Technology)

3:25 PM - 4:00 PM: Coffee Break

4:00 PM - 4:25 PM: SnAIPerf: A Reconfigurable Profiler for Multi-Granularity Analysis and Replay of CUDA Applications
Changhai Man, Tushar Krishna (Georgia Institute of Technology)

4:25 PM - 6:00 PM: Panel - Benjamin Lee (University of Pennsylvania), Tushar Krishna (Georgia Tech), Ankita Nayak (Gimlet Labs), Lishan Yang (George Mason University), Keren Zhou (George Mason University, OpenAI)

Call for Papers

We solicit both full papers (8-10 pages) and short/position papers (4-6 pages). Submissions are double-blinded. The page limit includes figures, tables, and appendices, but excludes references. Please use standard LaTeX or Word ACM templates. All submissions will need to be made via EasyChair (submission website: here). Each submission will be reviewed by at least three reviewers from the program committee. Papers will be reviewed for novelty, quality, technical strength, and relevance to the workshop. All accepted papers will be published here.

Important dates


Submission Deadline: 2/6/26
Acceptance Notification: 2/17/26
Camera Ready Submission Deadline: 3/6/26
Workshop date: 3/22/2026

All deadlines are at midnight anywhere on earth (AoE), and are firm.

Organization

Organizing Committee

  • Tom St. John, Gimlet Labs. (tomstjohn617@gmail.com)

  • Murali Emani, Argonne National Laboratory (memani@anl.gov)

  • Wenqian Dong, Oregon State University (wenqian.dong@oregonstate.edu>)

Program Committee

Qijing "Jenny" Huang (NVIDIA)
Gaurav Jain (xAI)
Gokcen Kestor (Barcelona Supercomputing Center)
Jiachen "Amber" Liu (Meta)
Xupeng Miao (Peking University)
Siddhisanket Raskar (Pacific Northwest National Laboratory)
Tanvi Sharma (NVIDIA)
Linghao Song (Yale)
Zishen Wan (Harvard)
Junqi Yin (ORNL)