Group
Current Members:
- Anton Lydike: Making accelerators more flexible
We're still growing :) See below for our current research and general topic ideas, and if interested in a PhD in compilers drop me an email.
Publications:
Jordi Armengol-Estapé, Rodrigo R.O. Rocha, Jackson Woodruff, Psquale Minervini, Michael F.P. O'Boyle, Forklift: An Extensible Neural Lifter, COLM 2024 PDF
Jackson Woodruff, Sam Ainsworth, Michael F.P. O'Boyle, Secco: Codesign for Resource Sharing in Regular Expression Accelerators, ASP-DAC 2024 PDF Code
Jordi Armengol-Estapé, Jackson Woodruff, Chris Cummins, Michael F.P. O'Boyle, SLaDe: A Portable Small Language Model Decompiler for Optimized Assemnbler, CGO 2024 Paper
Jackson Woodruff, Thomas Koehler, Alexander Brauckmann, Chris Cummins, Sam Ainsworth, Michael F.P. O'Boyle, Rewriting History: Repurposing Domain-Specific CGRAs, Arxiv PDF Code
José Wesley de Souza Magalhães, Jackson Woodruff, Elizabeth Polgreen, Michael F.P. O'Boyle, C2TACO: Lifting Tensor Code to TACO, GPCE 2023 Paper Code
Winner of GPCE Best Paper Award
Jackson Woodruff, David Schall, Michael F.P. O'Boyle, Christopher Woodruff, When does saving power save the planet?, HotCarbon 2023 Paper
Jackson Woodruff, Chris Cummins, Designing CGRAs with Deep Reinforcement Learning, DoSSA 2023 Paper Code
Pablo Antonio Martínez, Jackson Woodruff, Jordi Armengol-Estapé, Gregorio Bernabé, José Manuel García, Michael F.P. O'Boyle, Matching linear algebra and tensor code to specialized hardware accelerators, CC 2023 Preprint
Jackson Woodruff, Jordi Armengol-Estapé, Sam Ainsworth, Michael F.P. O'Boyle, Bind the Gap: Compiling Real Software to Hardware FFT Accelerators, PLDI 2022 Preprint Code Talk
Jordi Armengol-Estapé, Jackson Woodruff, Alexander Bruachkmann, José Wesley de Souza Magalhães, Michael F. P. O'Boyle, ExeBench: an ML-scale dataset of executable C functions, MAPS@PLDI 2022 Paper
Noa Zilberman, Andrew W Moore, Billy Cooper, Jackson Woodruff, Yuta Tokusashi, Pietro Bressana, Murali Ramanujam, Salvator Galea, NRG: A Netwrok Perspective on Applications' Performance, TMA 2021 PDF
Jackson Woodruff, Michael F. P. O'Boyle, New Regular Expressions on Old Accelerators, 58th Design Automation Conference, 2021. Preprint Code
Bruce Collie, Philip Ginsbach, Jackson Woodruff, Ajitha Rajan, Michael F. P. O'Boyle, M3: Semantic API Migrations, International Conference on Automated Software Engineering, 2020. PDF
Bruce Collie, Jackson Woodruff, Michael F.P. O'Boyle, Modeling black-box components with probabilistic synthesis, International Conference on Generative Programming: Concepts and Experiences, 2020. PDF Evaluation Code Winner of GPCE Best Paper Award
Jackson Woodruff, Andrew W Moore, Noa Zilberman, Measuring Burstiness in Data Center Applications, Buffers Workshop 2019. PDF
Jackson Woodruff, Murali Ramanumjam, Noa Zilberman, P4DNS: In-Network DNS, EuroP4 2019. PDF Code Video
(Also presented at ICIN 2020 and IETF Meeting 107 to the COIN sub-group)
Winner of the Xilinx Open Hardware Award 2020
Theses:
Jackson Woodruff, Acceleration for the many, not the few, PhD Thesis, 2024 PDF
Jackson Woodruff, Analyzing data center applications using high precision packet traces, Master's Dissertation, 2019. PDF
Jackson Woodruff, An Optimizing Compiler for ML, Undergraduate Dissertation, 2018. PDF Code
Other:
Jackson Woodruff, Michael F.P. O'Boyle, Rewriting in Modern Compilers, HiPEAC ACASES Summer School, 2023
Jackson Woodruff, Compiler Support for Hardware Accelerators, Arm Virtual Conference, 2021
Murali Ramanujam, Jackson Woodruff, Andrew W Moore and Noa Zilberman, "Reproducible Cloud Measurements", poster, Google Networking Research Summit, March 2019.
Jackson Woodruff, How to give a talk: annotated bibliography, unpublished 2019. PDF
Jackson Woodruff, How to give a talk (Talk at the Churchill CompSci Talk Series, 2019 Recording (Youtube))
I'm Jackson Woodruff
I'm a lecturer (assistant professor) at the University of Edinburgh working on compilers for hardware accelerators.
Currently Recruiting PhD Students, if interested send me an email.
Research Interests:
- How do we make hardware accelerators easy to use?
- How can we improve existing programming models (in particular, APIs and DSLs)?
- How do we make it easier to build compilers for hardware accelerators?
- What tools do programmers need to use hardware accelerators?
- How can we use compilers to co-design hardware?
- Can we leverage hardware accelerators to reduce carbon emissions?
J.C.Woodruff [at] sms.ed.ac.uk CV
Experience:
- Lecturer (Assistant Professor), University of Edinburgh, 2024-
- University of Cambridge, BA and MEng, 2015-2019
- University of Edinburgh, PhD + Postdoc, 2019-
- Meta AI, FAIR Labs Intern, 2022
- MIT, Visitor, 2023-24
- Arm Intern, 2017 and 2018
Awards:
- Best Paper Award, GPCE 2023
- Xilinx Open Hardware Award, 2020
- Best Paper Award, GPCE 2020
- Distinguished Talk Award, Churchill College Computer Science Talks Series, 2017 and 2018
- Magdalene College Prize for Computer Science, 2016, 2018, 2019
- Magdalene College Academic Scholarship, 2016-2017, 2018-2019, 2019-2020
- Top 50 A-Level Computer Science Student in the UK, 2015