Polyhedral Compilation Professors and Supervisors

Tobias Grosserمشاهده پروفایل

دانشیار

National and Kapodistrian University of AthensGreece

Tobias Grosser is an Associate Professor in the Department of Computer Science and Technology at the University of Cambridge. His research focuses on compiler technology, programming language design, and performance programming, with applications spanning hardware design, climate science, and quantum computing. He leads a research group developing innovative compiler frameworks and tools that bridge theoretical foundations with practical applications. Dr. Grosser completed his undergraduate studies in Computer Science at the University of Passau in Germany and pursued his PhD at École Normale Supérieure Paris as a Google PhD Fellow. Prior to joining Cambridge, he served as a Reader at the University of Edinburgh and held an Ambizione Fellowship at ETH Zurich. His research program centers on rethinking performance programming by re-connecting developers and compilers. He aims to make compilation more modular, predictable, automatic, and trustworthy while bringing open-source compiler innovation to increasingly diverse targets from GPUs to FPGAs and custom hardware. His work spans multiple domains including polyhedral compilation, constraint solving, quantum computing, and hardware design automation. Dr. Grosser is particularly interested in breaking down barriers between compilers and programmers by enabling their interaction through the programming language environment. His recent publications demonstrate a strong focus on compiler infrastructure development, particularly around the MLIR framework. He has pioneered work on Presburger arithmetic optimization with the FPL library, developed new intermediate representations for hardware description and quantum computing, and created tools for compiler education and prototyping like xDSL. His research shows a consistent theme of creating practical, high-performance compiler technologies that address real-world challenges across multiple domains. HiPEAC Technology Transfer Award 2021 for "Fast linear programming through transprecision computing on small and sparse data" OOPSLA 2021 Distinguished Paper Award for "FPL: Fast Presburger arithmetic through transprecision" Dr. Grosser actively mentors PhD students and postdoctoral researchers, currently supervising a team of over a dozen researchers working on various aspects of compiler technology. His group collaborates with industry partners including ARM and Xilinx, and maintains strong ties with the LLVM and MLIR open-source communities. He has secured funding for multiple research projects including work on verified compilation with Lean-MLIR, quantum compiler development, and hardware design automation. His research group operates at the intersection of multiple projects including Open-Source Electronic Design Automation, Seamless design of Smart Edge Processors, Lean-MLIR for verified compilation, FPL for fast Presburger arithmetic, and compilation frameworks for quantum computers. They maintain strong community engagement through regular Compiler Social events in Cambridge and active participation in LLVM developer meetings.

پروفایل رسمی دانشگاه

Saday Sadayappanمشاهده پروفایل

استاد

National and Kapodistrian University of AthensGreece

P. (Saday) Sadayappan is a Professor at the University of Utah's School of Computing, specializing in high-performance computing and compiler optimizations. His research focuses on performance optimization for parallel systems, particularly for tensor computations, sparse matrix operations, and machine learning workloads. He leads multiple NSF and DARPA-funded projects focused on GPU optimization, tensor computations, and scalable machine learning frameworks. Research Interests: Dr. Sadayappan's work spans compiler optimizations for high-performance systems, optimization of sparse/dense matrix/tensor computations, scalable machine learning, and algorithm-architecture co-design. His recent projects include developing performance-portable frameworks for tensor applications and optimizing data locality for scientific computing. Publication Trends: His recent publications (2020-2022) predominantly focus on GPU acceleration of machine learning workloads (especially CNNs), automated I/O complexity analysis, and optimization techniques for sparse matrix/tensor operations. Earlier work (2018-2019) established foundations in GPU code generation for tensor contractions and cache optimization. Awards and Honors: ACM SIGPLAN Most Influential PLDI Paper Award (2018) for A Practical Automatic Polyhedral Parallelizer and Locality Optimizer Active Grants and Projects: NSF: A Comprehensive Framework for Efficient, Scalable, and Performance-Portable Tensor Applications (2022-2027) NIH SBIR: Enabling next generation machine learning for large scale image analysis (2023-2025) NSF: AI Institute for Intelligent CyberInfrastructure (ICICLE) (2021-2026) NSF: Data Locality Optimization for Sparse Matrix/Tensor Computations (2020-2024) DARPA SBIR: Performance Portable Framework for Developing Graph Applications (2017-2022) Teaching: He currently teaches CS 4230/6230 (Parallel and High-Performance Computing) at the University of Utah.

پروفایل رسمی دانشگاه

Marc Moreno Mazaمشاهده پروفایل

استاد

Western UniversityCanada

Marc Moreno Maza is a Professor in the Computer Science and Applied Mathematics Departments at the University of Western Ontario, and a Principal Scientist at the Ontario Research Centre for Computer Algebra (ORCCA). His research focuses on applying computer science to mathematics, particularly polynomial system solving and algorithm design. Key areas include parallel computing, high-performance algebraic algorithms, and GPU acceleration. He leads projects funded by NSERC and industry partnerships, such as 'Hardware Acceleration Technologies for Polynomial Systems' and collaborative work with IBM and CAS Research. His research spans four directions: theoretical foundations of polynomial equations, efficient algorithm development, software implementation (e.g., RegularChains library in Maple), and applications to real-world challenges. He has delivered over 100 talks worldwide on topics like cylindrical algebraic decomposition, GPU-based polynomial arithmetic, and parametric system solving. His work emphasizes optimizing algorithms for modern architectures and leveraging parallelism. Notable contributions include the BPAS and CUMODP libraries for polynomial arithmetic, and the RegularChains library for semi-algebraic set computations. He collaborates internationally, with grants supporting both theoretical and applied research. His team addresses challenges in computational algebra, from theoretical breakthroughs to practical software tools.

پروفایل رسمی دانشگاه

Enea Zaffanellaمشاهده پروفایل

دانشیار

University of ParmaItaly

Enea Zaffanella is an Associate Professor in Computer Science at the University of Parma , where he has been affiliated since 2000. His academic career spans from Fellow Researcher to Assistant Professor, culminating in his current role since 2006. He teaches courses in Programming Methodologies , Compilers , and Foundations of Computer Science at both undergraduate and graduate levels. PhD in Computer Science from School of Computing, University of Leeds (2002) Laurea in Computer Science from University of Pisa (1993) Research focuses on Static Analysis and Software Verification using Abstract Interpretation . Key contributions include theoretical frameworks for Constraint Logic Programs analysis, Convex Polyhedra abstractions, and Widening Operators design. His work bridges formal theory with practical implementations like the Parma Polyhedra Library (PPL) and its successor PPLite . Recent publications explore Hybrid Systems verification, Data Science linting tools (Pyra), and EVM Bytecode analysis. He received the Radhia Cousot Young Researcher Best Paper Award in 2019. Collaborations include academic Research Projects (PRIN, ESPRIT) and industrial partnerships through BUGSENG srl .

پروفایل رسمی دانشگاه

David Monniauxمشاهده پروفایل

پژوهشگر

Max Planck Institute for Software SystemsGermany

David Monniaux is a senior researcher (directeur de recherche) at CNRS and an adjunct professor at École polytechnique. He works at VERIMAG, a computer science laboratory jointly operated by CNRS and the University of Grenoble. Dr. Monniaux obtained his PhD in 2001 from Université Paris Dauphine under Professor Patrick Cousot, with a dissertation on the static analysis of probabilistic programs by abstract interpretation. He later earned his habilitation in computer science in 2009 from Université Joseph Fourier, Grenoble, and also holds an agrégation in mathematics. Monniaux's research focuses on program verification, with particular emphasis on proving software correctness. His work spans theoretical foundations in computability theory and practical applications in safety-critical systems. He has made significant contributions to abstract interpretation, static analysis, and the verification of numerical properties in programs. His research bridges computer science theory with practical engineering challenges, particularly in the context of critical embedded systems where software failures can have severe consequences. His work connects to diverse fields including game theory, algebra, and convex optimization. His recent publications demonstrate a strong focus on improving the precision and efficiency of static analysis techniques. Key themes include polyhedral approximation, program analysis with local policy iteration, abstraction of arrays and maps, synthesis of ranking functions, and computing worst-case execution times. These works collectively advance the field of program verification by addressing challenges in handling nonlinear constraints, branching, and complex data structures while maintaining computational feasibility. Monniaux has supervised several students including Julien Henry, Alexis Fouilhé, George (Egor) Karpenkov, and Alexandre Maréchal (now at LIP6), with current students Hang Yu and Valentin Touzeau. He has led significant research projects including VERASCO (2012-2015), which aimed at integrating a static analyzer into the CompCert certified compiler, and STATOR (2012-2017), an ERC starting investigator grant exploring advanced techniques for automatic inference of program invariants. At VERIMAG, Monniaux is part of a vibrant research community focused on critical systems. His work connects with broader efforts in formal methods, with applications in aviation, automotive systems, and other safety-critical domains where software reliability is paramount.

پروفایل رسمی دانشگاه

P. Sadayappanمشاهده پروفایل

استاد

University of UtahUnited States

P. Sadayappan is a Professor at the School of Computing, University of Utah, specializing in high performance computing, compiler optimization, and scalable machine learning. His research focuses on developing efficient computational methods for scientific applications, particularly in the areas of sparse/dense matrix and tensor computations. His research interests include: Compiler Optimization for High Performance Computing Optimization of Sparse/Dense Matrix/Tensor Computations Scalable Machine Learning Algorithm-Architecture Co-Design Optimization Sadayappan's recent publications demonstrate a strong focus on tensor computations, GPU acceleration, and compiler optimizations for machine learning workloads. His work spans from fundamental compiler theory to practical implementations that improve performance across various architectures. A significant trend in his recent work involves the development of frameworks for efficient tensor operations, sparse matrix computations, and domain-specific code generation, with particular emphasis on performance portability across heterogeneous computing platforms. His notable scientific achievement includes receiving the ACM SIGPLAN Most Influential PLDI Paper Award in 2018 for his work on polyhedral compilation. Sadayappan has been principal investigator or co-investigator on numerous significant research grants, including: NSF award #2217154 (2022-2027): A Comprehensive Framework for Efficient, Scalable, and Performance-Portable Tensor Applications NSF award #2112606 (2021-2026): AI Institute for Intelligent CyberInfrastructure with Computational Learning in the Environment (ICICLE) NIH SBIR-Phase 2 (2023-2025): Enabling next generation machine learning for large scale image analysis NSF award #2009007 (2020-2024): Data Locality Optimization for Sparse Matrix/Tensor Computations DARPA SBIR-Phase 2 (2017-2022): Performance Portable Framework for Developing Graph Applications He teaches CS 4230/6230 (Parallel and High-Performance Computing) at the University of Utah and collaborates extensively with researchers across multiple institutions on projects involving computational chemistry, physics simulations, graph analytics, and machine learning.

پروفایل رسمی دانشگاه

Xavier Leroyمشاهده پروفایل

استاد

InriaFrance

Xavier Leroy is a Professor of Software Sciences at Collège de France and a senior computer scientist at Inria's Cambium research team. He is a member of the French Académie des sciences and internationally recognized for his work on functional programming, formal verification, and compiler design. Specializes in OCaml language development, Coq verification, and language-based security Leader of the CompCert formally-verified compiler project Received multiple ACM awards for software systems and programming languages achievement His research focuses on foundational aspects of programming languages, including type systems, concurrency, and algebraic effects. Recent work explores mechanized semantics, well-founded recursion, and persistent data structures. Key article trends include formal verification of compilers, functional data structure optimization, and effect handler implementation. Awards highlight community recognition of his 30+ year contributions to OCaml and CompCert. Notable Appointments: 2018–present: Professor, Collège de France (software sciences chair) Pre-2018: Senior Scientist, Inria

پروفایل رسمی دانشگاه

Michael Petterمشاهده پروفایل

مدرس

Technical University of MunichGermany

Michael Petter is a faculty member at the Technische Universität München (TUM), affiliated with the School of Computer Science II (Informatik 2) and the Department of Languages and Description Structures in Computer Science. He specializes in compiler design, static analysis, and formal verification, with a focus on developing tools like the Goblint framework for multi-threaded C programs. His teaching spans courses such as Compiler Construction, Program Optimization, and Functional Programming, reflecting his expertise in programming languages and software engineering. His research interests include static analysis techniques, concurrency control, and abstract interpretation frameworks. He has contributed to advancements in flow-sensitive analysis, context-sensitivity management, and precision recovery in program analysis. Recent work emphasizes improving static analysis tools for industrial benchmarks and enhancing their usability through incremental methods. Petter’s publications highlight innovations in numerical abstract domains (e.g., octagons), non-local jump misuse detection, and thread-modular analysis for multi-threaded systems. His involvement in projects like the DFG Research Training Group ConVeY underscores his commitment to advancing continuous verification in cyber-physical systems. He advises thesis projects on topics like quantum computing at compile time and program synthesis, fostering interdisciplinary collaboration. While actively engaged in academic events and editorial roles, he maintains a lab focused on cutting-edge compiler and static analysis research.

پروفایل رسمی دانشگاه

Richard Schulzeمشاهده پروفایل

پژوهشگر

University of MünsterGermany

Richard Schulze is a Researcher at the University of Münster, contributing to projects such as SkelCL, PACXX, and dOpenCL. His work focuses on parallel computing, compiler optimization, and auto-tuning frameworks for high-performance and distributed systems. He explores portable code generation for heterogeneous architectures using Multi-Dimensional Homomorphisms (MDH) and develops abstractions for OpenCL/CUDA programming. His research interests include advancing scheduling languages and systematic composition models, alongside probabilistic data linkage techniques. Recent publications emphasize auto-tuning methodologies for Python and interdependent parallel program parameters. Publications since 2018 highlight contributions to portable compiler design, performance optimization, and cross-platform parallelism. No scientific awards are explicitly mentioned. Consultation hours are by appointment, and he is affiliated with the university's computer science research groups.

پروفایل رسمی دانشگاه

Tobias Grosserمشاهده پروفایل

دانشیار

University of CambridgeUnited Kingdom

Tobias Grosser is an Associate Professor in the Department of Computer Science and Technology at the University of Cambridge. His research focuses on rethinking performance programming by bridging the gap between developers and compilers. He holds a PhD from École Normale Supérieure Paris and has held positions including Reader at the University of Edinburgh and Ambizione Fellow at ETH Zurich. His research interests span compilers, programming language design, static/dynamic analysis, and the integration of machine learning into compiler development. He emphasizes making compilation more modular, automatic, and trustworthy, with applications in quantum computing, climate science, and open-source hardware. Key projects include xDSL (a Python-native compiler framework), LoopOpt, and the Open Earth Compiler for climate simulations. Recent publications highlight advancements in multi-level intermediate representations (IR), formal verification in MLIR, and performance optimization for GPUs and FPGAs. His work often addresses barriers between programmers and compilers, aiming for intuitive collaboration between developers and automated systems. Tobias mentors a dynamic team of PhD students, postdocs, and researchers, including notable contributors like Siddharth Bhat, Arjun Pitchanathan, and Mathieu Fehr. His lab focuses on compiler toolchains for domain-specific hardware accelerators, quantum computing ecosystems, and verified compilation techniques.

پروفایل رسمی دانشگاه

Fredrik Kjolstadمشاهده پروفایل

استادیار

National and Kapodistrian University of AthensGreece

Fredrik Kjolstad is an Assistant Professor in the Department of Computer Science at Stanford University, specializing in compilers and programming models for sparse computing and performance engineering. His research focuses on separating algorithms from data representations to enable portable applications across diverse hardware platforms. His research interests span compilers, programming models, performance engineering, and computer architecture, with particular emphasis on sparse tensor algebra, compiler design for heterogeneous systems, and high-performance computing. He has pioneered frameworks like TACO, Simit, and Distal that enable efficient sparse computations across CPUs, GPUs, and specialized accelerators. Dr. Kjolstad's publications demonstrate expertise in compiler optimization techniques for sparse data structures, tensor algebra, and distributed systems. His work consistently addresses the challenge of bridging high-level programming abstractions with efficient hardware execution across diverse architectures. MIT EECS First Place George M. Sprowls PhD Thesis Award NSF CAREER Award Rosing Award Adobe Fellowship Google Research Scholarship Best Paper Awards at EuroMPI 2013, OOPSLA 2017, and OOPSLA 2021 ISCA Distinguished Artifact Award PLDI and OOPSLA Distinguished Paper Awards He advises multiple PhD students including James Dong, Olivia Hsu, and Rohan Yadav, while leading research on compiler technologies that have received significant grant support. His group develops practical tools like the TACO compiler and Legate Sparse that are used in both academic and industrial settings. Current projects focus on programmable accelerators for sparse tensor algebra, distributed sparse computing, and compiler support for emerging hardware architectures.

پروفایل رسمی دانشگاه

Clément QUINTONمشاهده پروفایل

دانشیار

University of LilleFrance

Clément QUINTON is an Associate Professor at the University of Lille, affiliated with both Inria and the CRIStAL laboratory (Centre de Recherche en Informatique, Signal et Automatique de Lille). He is an elected member of the laboratory council and maintains offices in both the Inria building (Room B312) and the M3 building (Room 224) at the Cité Scientifique campus. Dr. QUINTON is a member of the Spirals research team and holds the HDR (Habilitation à Diriger des Recherches), a French academic qualification that allows him to supervise PhD students. Dr. QUINTON's research spans several areas within computer science, with a primary focus on parallel computing, polyhedral models, and hardware acceleration. His work bridges theoretical computer science with practical applications in embedded systems and biomedical monitoring. He has made significant contributions to the fields of: Polyhedral compilation and loop optimization techniques High-level synthesis for FPGA-based hardware acceleration Systolic array design and parallel architectures Disruption-tolerant wireless sensor networks for biomedical applications Sustainable software engineering and cloud configuration Application of Large Language Models to software development Analysis of Dr. QUINTON's publication history reveals a consistent research trajectory that began with foundational work in parallel algorithms and systolic arrays, evolving toward more contemporary applications involving FPGA acceleration, polyhedral compilation, and biomedical sensor networks. His recent work shows increasing interest in sustainable computing, cloud architectures, and the application of Large Language Models to software engineering challenges. This evolution demonstrates his ability to adapt theoretical computer science concepts to address emerging technological challenges. Dr. QUINTON has supervised numerous PhD students whose research aligns with his expertise: Virginie Amand: Sustainable software services using large-scale models Alexandre Bonvoisin: Frugal software architectures for cloud-native microservices Tristan Coignion: Energy impact of Large Language Models for code Edouard Guegain: Software optimization through configuration (defended September 2023) Brell Péclard Sanwouo Chekam: Automatic detection and correction of side-channel vulnerabilities in cryptographic libraries Nada Zine: Complex and self-adaptive software systems Maxime Huyghe: Sustainable software services development based on Language Models (LLM) As a member of the Spirals research team at Inria and CRIStAL, Dr. QUINTON contributes to a collaborative environment focused on software engineering, system architecture, and sustainable computing. His work bridges theoretical computer science with practical applications in healthcare monitoring, cloud computing, and energy-efficient software development.

پروفایل رسمی دانشگاه