Tyler Sorensen

About

I am a Principal Researcher in the RiSE group at Microsoft Research, and an Associate Professor (on leave) at UC Santa Cruz. My research focuses on concurrency: programming, modeling, testing, and architecture, with the goal of enabling correct and efficient applications on emerging architectures. I explore these ideas primarily through GPGPU programming.

Previously, I was a postdoc at Princeton with Margaret Martonosi and received my PhD from Imperial College London advised by Alastair Donaldson.

PhD Students

• Reese Levine - Started Fall 2020
• Yanwen Xu - Started Spring 2021
• Rithik Sharma - Started Fall 2022
• Zheyuan Chen - Started Fall 2024

Agentic Workflows

I am experimenting with agentic workflows. Enjoy!

Did You Know? (daily fact mined from my papers)

Did you know? Traditional simultaneous multithreading (SMT) designs must choose between out-of-order (OoO) threads (for instruction-level parallelism) or in-order (InO) threads (for thread-level parallelism) — but SHADOW is the first CPU architecture to run both kinds of threads concurrently on the same core, dynamically stealing work between them. On memory-bound sparse workloads like those found in deep learning and graph processing, SHADOW achieves up to 3.16× speedup and 1.33× average improvement over a standard OoO CPU, with only 1% area and power overhead. (from: SHADOW: Simultaneous Multi-Threading Architecture with Asymmetric Threads)

Research Idea (daily idea mined from my papers and related works on arXiv)

Placeholder: check back soon — an agent will populate this with a research idea inspired by my papers and recent arXiv publications.

Publications

All papers are provided in full markdown and summary markdown formats to make them easier for agents and LLMs to process. See the [Markdown Full] and [Markdown Summary] links next to each paper, or browse the combined index.

Papers

Workshop Papers

Theses

Teaching

I have taught classes at UCSC on compilers and parallel programming.

• CSE 280O-01: Language Systems and Data (LSD) Seminar (co-organized with Lindsey Kuper) - Fall 2023
• CSE 211: Graduate Compiler Design - Fall 2023
• CSE 113: Concurrent and Parallel Programming - Winter 2023
• CSE 110A: Fundamentals of Compiler Design I - Spring 2023

Alumni

I've had the privilege of working with the following students at UCSC who have gone on to really great things!

• Devon McKee - MS → Samsung
• Sanya Srivastava - MS → PhD student at Duke
• Christian Lei - MS → Microsoft
• Kiefer Selmon - MS → Nvidia
• Mingun Cho - UG → PhD student at UC Davis
• Chris Liu - UG → PhD student at UCLA
• Naomi Rehman - UG → PhD student at UC Santa Barbara
• Sean Siddens - UG → PhD student at University of Washington
• Ananthajit Srikanth - UG → PhD student at EPFL

Acknowledgements

I would like to thank various organizations for supporting my work, either through employment, contracting, or research funding: Microsoft, Trail of Bits, NSF, Google, DARPA, and the Khronos Group.