PeopleofSuperintelligence

Researcher

Milad Hashemi

Milad Hashemi is a researcher whose work applies machine learning to computer systems, from neural memory prefetching to AI-driven code optimization. They studied electrical engineering at the University of Washington and UT Austin, then completed an electrical and computer engineering PhD at UT Austin in 2016. Their work spans neural models for numerical subroutines, amortized sampling for energy-based models, and deployed systems that tune memory access and optimize production code. A coauthored discrete-sampling paper received an ICML 2021 Outstanding Paper Honorable Mention. They are a Principal Scientist at Google DeepMind.

Badges

Timeline

  1. 2005–2016

    Three engineering degrees

    They earn a B.S. in electrical engineering at the University of Washington (2005–2009) and an M.S. in electrical and computer engineering at UT Austin (2009–2011). They complete a PhD in electrical and computer engineering at UT Austin in 2016 as a member of the HPS research group, advised by Professor Yale Patt.

  2. 2018

    Neural memory prefetching

    At ICML 2018, they coauthor “Learning Memory Access Patterns,” which uses recurrent neural networks to predict memory access and build data prefetchers. The paper reports improved precision and recall on benchmark datasets.

  3. 2020

    Neural execution engines

    At NeurIPS 2020, they coauthor “Neural Execution Engines: Learning to Execute Subroutines,” a study of how transformers learn numerical subroutines such as sorting, shortest paths and minimum spanning trees. The paper proposes learned conditional masking and binary number embeddings to help models generalize beyond their training examples.

  4. 2021

    Amortized energy-model sampling

    At ICLR 2021, they coauthor “No MCMC for Me,” which trains energy-based models with an entropy-regularized generator that amortizes MCMC sampling. The approach extends Joint Energy Models to semi-supervised classification on tabular data.

  5. 2021

    An ICML honorable mention

    In 2021, they and four coauthors receive an ICML Outstanding Paper Honorable Mention for “Oops I Took A Gradient,” a scalable sampler for discrete probability models. The method uses likelihood gradients to propose updates in a Metropolis-Hastings sampler.

  6. 2024

    Prefetching at scale

    In 2024, they coauthor “Limoncello,” a system that adjusts data prefetching for heavily used computing systems. It disables hardware prefetching when memory bandwidth use is high and uses targeted software prefetching to reduce cache misses; deployment in a hyperscale system raises throughput by 10% and lowers memory latency by 15%.

  7. As of October 2026

    Principal scientist at Google DeepMind

    They are a Principal Scientist at Google DeepMind. At OSDI 2026, they coauthor “ECO: An AI-Driven Code Efficiency Optimizer for Warehouse Scale Computers,” which automatically modifies production source code to improve performance. The system lands more than 6,400 commits and changes over 25,000 lines of production code at Google.

Career

  1. NowGoogle DeepMindPrincipal Scientist

Education

  1. 2016University of Texas at AustinPhD
  2. 2011University of Texas at AustinM.S.
  3. 2009University of WashingtonB.S.

From the biography.

  1. 16% of AI career moves in the 2020s went from a university to a companyIncludes Milad Hashemi. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  2. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  3. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.

Sources and credits

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

  1. Milad Hashemi sites.google.com. Checked 21 October 2026.
  2. Milad Hashemi hps.ece.utexas.edu. Checked 21 October 2026.
  3. Learning Memory Access Patterns research.google. Checked 21 October 2026.
  4. Neural Execution Engines: Learning to Execute Subroutines research.google. Checked 21 October 2026.
  5. No MCMC for Me: Amortized Samplers for Fast and Stable Training of Energy-Based Models research.google. Checked 21 October 2026.
  6. ICML 2021 Awards icml.cc. Checked 21 October 2026.
  7. Limoncello: Prefetchers for Scale research.google. Checked 21 October 2026.
  8. ECO: An AI-Driven Code Efficiency Optimizer for Warehouse Scale Computers (Operational Systems) usenix.org. Checked 21 October 2026.