Reese Levine

profile.png
Santa Cruz, CA

I am a fourth year PhD candidate and an NDSEG Fellow at UC Santa Cruz, where I work with Tyler Sorensen and am generally affiliated with the LSD Lab.

My research interests are in parallel and concurrent programming, specifically how multi-core and distributed systems are implemented. My current work focuses on GPU memory models, designing and evaluating techniques to test the conformance of compilers and hardware to specifications. I am also working on ways to exploit weak memory behaviors to increase applications’ performance, and to test and improve the safety properties of GPU programming models in the face of data races or other undefined behavior.

For a demo of our work on testing GPU memory models, check out our testing website GPUHarbor, which is implemented using the WebGPU framework. If you want to contribute to our research, head to the Tune/Conform page and follow the instructions there to run tests and submit experimental data about your GPU’s shared memory behaviors.

News

  • I presented our work on testing the Vulkan memory model at Vulkanised 2024.
  • I gave a talk on evolving weak memory models for evolving architectures at the Future of Weak Memory workshop at POPL 2024.
  • I presented our work on testing memory consistency at Imperial College London, University of Kent, Cambridge University, and Bristol University [January 2024].
  • I returned for another summer internship at Apple on their GPU Platform Architecture team [June-September 2023].
  • I presented GPUHarbor at ISSTA 2023, where it won a Distinguished Artifact Award [June 2023].
  • Our experience paper on running a large scale GPU memory model study, nicknamed GPUHarbor, was accepted to ISSTA 2023 [April 2023].
  • I was honored to be awarded an NDSEG Fellowship [April 2023].
  • MC Mutants won both a Distinguished Paper Award and a Distinguished Artifact Award at ASPLOS 2023 [March 2023]. Check out this article from UCSC with more details about our work.
  • My proposal Probabilistic Memory Consistency Specifications was accepted to YArch’23, a workshop at ASPLOS 2023 [February 2023].
  • I passed my qualifying examination and advanced to candidacy, presenting my thesis proposal Testing and Improving Memory Consistency Specifications [January 2023].
  • I presented the MC Mutants work at the Khronos Group’s F2F in Phoenix, AZ [October 2022].
  • Our work on evaluating testing techniques for memory models, MC Mutants, was accepted to ASPLOS 2023 [September 2022].
  • I interned with Apple’s GPU Platform Architecture team [July-September 2022].
  • I presented our ongoing work on testing memory models at UCSC’s LSD seminar [January 2022].

Teaching

  • TA, Parallel Programming, UC Santa Cruz, Spring 2021/Winter 2022/Winter 2023
  • TA, Computer Architecture, UC Berkeley Summer/Fall 2016

Personal

When I’m not thinking about computers (and sometimes when I am), I prefer to spend my time outside, including cycling, climbing, skiing, surfing Santa Cruz’s many breaks, and playing on the beach with our dog Hunter.