Phitchaya Mangpo Phothilimthana

Blog Posts

• Advancements in machine learning for machine learning
  Phitchaya Mangpo Phothilimthana and Bryan Perozzi.
  Google Research, 2023. (post)

• Google Research, 2022 & beyond: ML & computer systems
  Phitchaya Mangpo Phothilimthana and Adam Paszke.
  Google Research, 2023. (post)

Peer-Reviewed Papers

• Accelerating Retrieval-Augmented Language Model Serving with Speculation
  Zhihao Zhang, Alan Zhu, Lijie Yang, Yihua Xu, Lanting Li, Phitchaya Mangpo Phothilimthana, Zhihao Jia.
  International Conference on Machine Learning (ICML), 2024.
  (PDF)

• Thesios: Synthesizing Accurate Counterfactual I/O Traces from I/O Samples.
  Phitchaya Mangpo Phothilimthana, Saurabh Kadekodi, Soroush Ghodrati, Selene Moon, Martin Maas.
  Architectural Support for Programming Language and Operating Systems (ASPLOS), 2024.
  (PDF, TRACES, slides)

• BRAINSTORM: Supercharging Innovation with AI-Driven Ideation.
  Gemini, Deniz Altınbüken, Martin Maas, Phitchaya Mangpo Phothilimthana.
  ASPLOS Wild and Crazy Ideas (WACI), 2024.
  (PDF)

• TpuGraphs: A Performance Prediction Dataset on Large Tensor Computational Graphs
  Phitchaya Mangpo Phothilimthana, Sami Abu-El-Haija, Kaidi Cao, Bahare Fatemi, Mike Burrows, Charith Mendis, and Bryan Perozzi.
  Advances in Neural Information Processing Systems (NeurIPS), 2023.
  (PDF, slides)

• Large Graph Property Prediction via Graph Segment Training
  Kaidi Cao, Phitchaya Mangpo Phothilimthana, Sami Abu-El-Haija, Dustin Zelle, Yanqi Zhou, Charith Mendis, Jure Leskovec, and Bryan Perozzi.
  Advances in Neural Information Processing Systems (NeurIPS), 2023.
  (PDF)

• Neural Architecture Search Using Property Guided Synthesis
  Charles Jin, Phitchaya Mangpo Phothilimthana, and Sudip Roy
  Object-oriented Programming, Systems, Languages, and Applications (OOPSLA), 2022.
  (PDF)

• GRANITE: A Graph Neural Network Model for Basic Block Throughput Estimation
  Ondrej Sykora, Phitchaya Mangpo Phothilimthana, Charith Mendis, and Amir Yazdanbakhsh
  IEEE International Symposium on Workload Characterization (IISWC), 2022.
  (PDF)

• A Transferable Approach for Partitioning Machine Learning Models on Multi-Chip-Modules
  Xinfeng Xie, Prakash Prabhu, Ulysse Beaugnon, Phitchaya Mangpo Phothilimthana, Sudip Roy, Azalia Mirhoseini, Eugene Brevdo, James Laudon, and Yanqi Zhou
  Conference on Machine Learning and Systems (MLSys), 2022.
  (PDF)

• A Flexible Approach to Autotuning Multi-Pass Machine Learning Compilers
  Phitchaya Mangpo Phothilimthana et al.
  Conference on Parallel Architectures and Compilation Techniques (PACT), 2021.
  (PDF, slides)

• Equality Saturation for Tensor Graph Superoptimization
  Yichen Yang, Phitchaya Mangpo Phothilimthana, Yisu Remy Wang, Max Willsey, Sudip Roy, and Jacques Pienaar
  Conference on Machine Learning and Systems (MLSys), 2021.
  (PDF)

• A Learned Performance Model for Tensor Processing Units
  Samuel J. Kaufman, Phitchaya Mangpo Phothilimthana, Yanqi Zhou, Charith Mendis, Sudip Roy, Amit Sabne, and Mike Burrows.
  Conference on Machine Learning and Systems (MLSys), 2021.
  (PDF)

• Transferable Graph Optimizers for ML Compilers
  Yanqi Zhou, Sudip Roy, Amirali Abdolrashidi, Daniel Wong, Peter Ma, Qiumin Xu, Hanxiao Liu, Phitchaya Mangpo Phothilimthana, Shen Wang, Anna Goldie, Azalia Mirhoseini, and James Laudon.
  Advances in Neural Information Processing Systems (NeurIPS), 2020.
  (PDF)

• Learning Local Advantage Functions for Generalizable Graph Optimizations
  Yifan Wu, Yanqi Zhou, Phitchaya Mangpo Phothilimthana, Hanxiao Liu, Sudip Roy, and Azalia Mirhoseini.
  ML for Systems Workshop at Neural Information Processing Systems (NeurIPS), 2020.
  (PDF)

• Learned TPU Cost Model for XLA Tensor Programs
  Samuel J. Kaufman, Phitchaya Mangpo Phothilimthana, and Mike Burrows.
  ML for Systems Workshop at Neural Information Processing Systems (NeurIPS), 2019.
  (PDF)

• E3: Energy-Efficient Microservices on SmartNIC-Accelerated Servers
  Ming Liu, Simon Peter, Arvind Krishnamurthy, and Phitchaya Mangpo Phothilimthana.
  USENIX Annual Technical Conference (ATC), 2019.
  (PDF)

• Swizzle Inventor: Data Movement Synthesis for GPU Kernels.
  Phitchaya Mangpo Phothilimthana, Archibald Samuel Elliott, An Wang, Abhinav Jangda, Bastian Hagedorn, Henrik Barthels, Samuel J. Kaufman, Vinod Grover, Emina Torlak, and Rastislav Bodik.
  Architectural Support for Programming Language and Operating Systems (ASPLOS), 2019.
  (PDF, slides)

• Floem: A Programming System for NIC-Accelerated Network Applications.
  Phitchaya Mangpo Phothilimthana, Ming Liu, Antoine Kaufmann, Simon Peter, Rastislav Bodik, and Thomas Anderson.
  USENIX Symposium on Operating Systems Design and Implementation (OSDI), 2018. (PDF, slides)

• Data-Driven Synthesis of Full Probabilistic Programs.
  Sarah Chasins and Phitchaya Mangpo Phothilimthana.
  Conference on Computer-Aided Verification (CAV), 2017. (PDF)

• High-Coverage Hint Generation for Massive Courses: Do Automated Hints Help CS1 Students?
  Phitchaya Mangpo Phothilimthana and Sumukh Sridhara.
  Conference on Innovation and Technology in Computer Science Education (ITiCSE), 2017. (PDF)

• Domain-Specific Symbolic Compilation
  Rastislav Bodik, Kartik Chandra, Phitchaya Phothilimthana, and Nathaniel Yazdani.
  Summit oN Advances in Programming Languages (SNAPL), 2017. (PDF)

• Short and Simple Cycle Separators in Planar Graphs.
  Eli Fox-Epstein, Shay Mozes, Phitchaya Mangpo Phothilimthana, and Christian Sommer.
  Journal of Experimental Algorithmics (JEA), 2016. (PDF)

• Compiling a Gesture Recognition Application for a Low-Power Spatial Architecture.
  Phitchaya Mangpo Phothilimthana, Michael Schuldt, and Rastislav Bodik.
  Conference on Languages, Compilers, Tools and Theory for Embedded Systems (LCTES), 2016. (PDF, slides)

• Scaling up Superoptimization.
  Phitchaya Mangpo Phothilimthana, Aditya Thakur, Rastislav Bodik, and Dinakar Dhurjati.
  Architectural Support for Programming Language and Operating Systems (ASPLOS), 2016. (PDF, slides)

• GreenThumb: Superoptimizer Construction Framework.
  Tool Demonstration Paper.
  Phitchaya Mangpo Phothilimthana, Aditya Thakur, Rastislav Bodik, and Dinakar Dhurjati.
  Conference on Compiler Construction (CC), 2016. (PDF, demo)

• Dicer: A Framework for Controlled, Large-Scale Web Experiments.
  Sarah Chasins and Phitchaya Mangpo Phothilimthana Temporal Web Analytics Workshop (TempWeb), 2015. (PDF)

• Chlorophyll: Synthesis-Aided Compiler for Low-Power Spatial Architectures.
  Phitchaya Mangpo Phothilimthana, Tikhon Jelvis, Rohin Shah, Nishant Totla, Sarah Chasins, and Rastislav Bodik.
  Programming Languages Design and Implementation (PLDI), 2014. (PDF, slides)

• A Comparison of Error Metrics for Learning Model Parameters in Bayesian Knowledge Tracing.
  Phitchaya Mangpo Phothilimthana, Asif Dhanani, Seung Yeon Lee, and Zachary Pardos.
  Bayesian Knowledge Tracing 20 Years Workshop (BKT20y), 2014. (PDF, poster)

• Communication-Minimizing 2D Convolution in GPU Registers.
  Forrest Iandola, David Sheffield, Michael Anderson, Phitchaya Mangpo Phothilimthana, and Kurt Keutzer.
  International Conference on Image Processing (ICIP), 2013. (PDF)

• Portable Performance on Heterogeneous Architectures.
  Phitchaya Mangpo Phothilimthana, Jason Ansel, Jonathan Ragan-Kelley, and Saman Amarasinghe.
  Architectural Support for Programming Language and Operating Systems (ASPLOS), 2013. (PDF, slides)

• Short and Simple Cycle Separators in Planar Graphs.
  Eli Fox-Epstein, Shay Mozes, Phitchaya Mangpo Phothilimthana, and Christian Sommer.
  Algorithm Engineering and Experiments (ALENEX), 2013. (PDF)

• MRSI module and SIVIC interface.
  Bjoern Menze, Mangpo Phothilimthana, Olson Beck, Jason Crane, and Polina Golland.
  National Alliance for Medical Image s Summer Project week (NAMIC), 2010. (URL)

arXiv & Technical Reports

• Optimizing Memory Mapping Using Deep Reinforcement Learning.
  Pengming Wang, Mikita Sazanovich, Berkin Ilbeyi, Phitchaya Mangpo Phothilimthana, Manish Purohit, Han Yang Tay, Ngân Vũ, Miaosen Wang, Cosmin Paduraru, Edouard Leurent, Anton Zhernov, Po-Sen Huang, Julian Schrittwieser, Thomas Hubert, Robert Tung, Paula Kurylowicz, Kieran Milan, Oriol Vinyals, Daniel J. Mankowitz.
  arXiv, 2023. (PDF)

• Dueling Metrics: Choosing the Appropriate Error Metric for Models of Cognition in the Learning Analytics Field.
  Phitchaya Phothilimthana, Seung Yeon Lee, and Zachary Pardos.
  University of California, Berkeley, 2018. (PDF)

Selected Invited Talks

• Machine Learning for Machine Learning Compilers.
  Stanford MLSys Seminar, 2023. (slides, recording)

• ML for Autotuning Production ML Compilers.
  ML for Systems Workshop (NeurIPS), 2021. (slides)

• Self-Evolving Compilers.
  Conference of Program Synthesis, San Franciscos, 2019. (slides)

• High-Coverage Hint Generation for Racket Programming Assignments.
  RacketCon, Seattle, 2017. (slides)

Theses

• Programming Abstractions and Synthesis-Aided Compilation for Emerging Computing Platforms.
  PhD Thesis.
  University of California, Berkeley, 2018. (PDF)

• Chlorophyll: Synthesis-Aided Compiler for Low-Power Spatial Architectures.
  Master Thesis.
  University of California, Berkeley, 2015. (PDF)