Researcher

Madanlal Musuvathi

Also known as Madan Musuvathi

Madanlal Musuvathi is a computer scientist at Microsoft Research whose tools hunt down concurrency bugs and whose recent systems handle communication between GPUs for AI workloads. After a computer science degree at IIT Madras, he earned a Stanford doctorate in 2004 working with David Dill and Dawson Engler, built the CMC model checker, and joined Microsoft Research. There he presented CHESS in 2008 with Shaz Qadeer and Thomas Ball, a tool that takes control of thread scheduling to reproduce Heisenbugs, and his 2019 thread-safety work won a SOSP Best Paper Award. His SCCL and MSCCL++ work targeted deep-learning training and Microsoft Azure AI services, and he shared the 2023 CAV Award for context-bounded analysis. Named an ACM Fellow in January 2026, he now works there as a Partner Research Manager.

Works
213
Citations
6,947
h-index
42

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. By 2004

    IIT Madras and Stanford

    He earns a B.Tech in computer science at IIT Madras, then an M.S. and a Ph.D. at Stanford, where he works with David L. Dill and Dawson Engler and completes the doctorate in 2004. In 2002, he presents CMC at OSDI, a model checker that checks C and C++ implementations directly.

  2. 2004

    Joining Microsoft Research

    He joins Microsoft Research in 2004. His paper with Junfeng Yang, Paul Twohey and Dawson Engler, "Using Model Checking to Find Serious File System Errors," wins a Best Paper award at OSDI 2004.

  3. 2008–2010

    The CHESS concurrency tester

    In 2008, he presents CHESS at OSDI with Microsoft Research colleagues Shaz Qadeer and Thomas Ball; the tool takes control of thread scheduling to find and reproduce Heisenbugs in concurrent programs, and Microsoft testers use it daily. In March 2010, he coauthors an ASPLOS paper with Sebastian Burckhardt and Pravesh Kothari on a randomized scheduler that finds buggy thread schedules more effectively than stress testing.

  4. 2019

    RiSE and a SOSP best paper

    In 2019, he manages the Research in Software Engineering (RiSE) group at Microsoft Research. His paper with Guangpu Li, Shan Lu, Suman Nath and Rohan Padhye, "Efficient and Scalable Thread-Safety Violation Detection," catches thousands of concurrency bugs and wins a SOSP 2019 Best Paper Award.

  5. 2020–2025

    SCCL and MSCCL++ for AI

    In August 2020, he coauthors SCCL, a system that synthesizes collective communication algorithms tailored to a hardware topology, the communication step that bottlenecks data-parallel deep-learning training; the paper wins the PPoPP 2021 Best Paper Award. In April 2025, he coauthors MSCCL++, a design method for portable GPU communication kernels that runs in production in multiple Microsoft Azure AI services and is adopted by AMD's RCCL library.

  6. 2023

    A CAV Award

    He shares the 2023 CAV Award with Akash Lal, Shaz Qadeer, Jakob Rehof and Thomas Reps for introducing context-bounded analysis and applying it to the systematic testing of concurrent programs.

  7. January 2026

    Named an ACM Fellow

    In January 2026, ACM names him a Fellow for the development of methods in concurrency verification and testing, and machine learning systems design.

  8. As of October 2026

    Partner research manager at Microsoft

    He works as a Partner Research Manager at Microsoft Research. In 2026, he serves as program co-chair of the Principles and Practice of Parallel Programming (PPoPP) conference.

Career

  1. Since 2004Microsoft ResearchPartner Research Manager, Manager
  2. 2005–2026Microsoft

Education

  1. 2004Stanford UniversityM.S.
  2. IIT MadrasB.Tech

Most-cited works

  1. Finding and Reproducing Heisenbugs in Concurrent Programs2008 · 542 citations
  2. Iterative context bounding for systematic testing of multithreaded programs2007 · Proceedings of the ACM SIGPLAN Conference on Programming Language Design and Implementation · 484 citations
  3. CMC2002 · ACM SIGOPS Operating Systems Review · 361 citationsWith Dawson Engler
  4. Using model checking to find serious file system errors2006 · ACM Transactions on Computer Systems · 310 citationsWith Dawson Engler
  5. LiteRace2009 · Proceedings of the ACM SIGPLAN Conference on Programming Language Design and Implementation · 245 citations

Sources and credits

Photo: We have no published photo of Madanlal Musuvathi yet. Send us a free-licensed photo.

  1. Madan Musuvathi - SPLASH 2014 2014.splashcon.org. Checked 20 October 2026.
  2. Madan Musuvathi cs.columbia.edu. Checked 20 October 2026.
  3. CMC: A Pragmatic Approach to Model Checking Real Code | USENIX usenix.org. Checked 20 October 2026.
  4. USENIX: OSDI '04 Technical Sessions usenix.org. Checked 20 October 2026.
  5. Finding and Reproducing Heisenbugs in Concurrent Programs | USENIX usenix.org. Checked 20 October 2026.
  6. Finding and Reproducing Heisenbugs in Concurrent Programs static.usenix.org. Checked 20 October 2026.
  7. A Randomized Scheduler with Probabilistic Guarantees of Finding Bugs - Microsoft Research microsoft.com. Checked 20 October 2026.
  8. Madan Musuvathi lcpc19.cc.gatech.edu. Checked 20 October 2026.
  9. SIGOPS 2019 Awards – ACM SIGOPS sigops.org. Checked 20 October 2026.
  10. New Concurrency Bug Approach From Shan Lu’s Group Wins Best Paper at SOSP – Department of Computer Science computerscience.uchicago.edu. Checked 20 October 2026.
  11. [2008.08708] Synthesizing Optimal Collective Algorithms arxiv.org. Checked 20 October 2026.
  12. PPoPP 2021 ppopp21.sigplan.org. Checked 20 October 2026.
  13. [2504.09014] MSCCL++: Rethinking GPU Communication Abstractions for AI Inference arxiv.org. Checked 20 October 2026.
  14. CAV Award | International Conference on Computer-Aided Verification i-cav.org. Checked 20 October 2026.
  15. Research in Software Engineering (RiSE) - Microsoft Research: News And Awards microsoft.com. Checked 20 October 2026.
  16. CRA Congratulates New AAAI, ACM, and IEEE-CS Fellows - CRN cra.org. Checked 20 October 2026.
  17. Madan Musuvathi - CGO 2026 2026.cgo.org. Checked 20 October 2026.