Talks

Keynote addresses, distinguished lectures, and invited seminars given at universities, national laboratories, research institutes, and conferences since 2001. Conference and workshop papers are listed on the publications page.

2025

• SWEET: A Hardware & Software Stack for Wrist-to-Cloud IntelligenceNSF CSR Principal Investigators Meeting · Boca Raton, FL
• SWEET: System Architectures and Programming Abstractions for Resilient and Efficient Edge AIDiscover-US Webinar, European Commission · virtual

2024

• Sustainable AI from the Viewpoint of an AI SkepticDistinguished Speaker Seminar, Department of Computer Engineering and Informatics · virtual

2021

• Programming Models for High-Performance Computing: Key Drivers and Barriers to AdoptionInvited talk, First NI-HPC Users Conference · virtual
• Why Should We Still be Concerned about Energy Management in System Software?Invited talk, 12th International Green and Sustainable Computing Conference · virtual

2018

• AVA: Accelerator Virtualization AnywhereKeynote13th Workshop on Virtualization in High-Performance Cloud Computing (VHPC)
• From Approximate to Significance-Driven to Transprecision Computing: Challenges and OpportunitiesDistinguished Speaker Seminar, Department of Computational Science, ETH Zurich
• Virtualizing Any Accelerator AnywhereDistinguished Speaker Seminar, Department of Computer Science, College of William & Mary

2016

• Energy Efficient Computing using Computational Significance AbstractionsKeynoteUK-China Workshop on Shaping Low Carbon Energy Future · Belfast, UK
• Using Computational Significance and Resilience in System Software StacksKeynoteFirst Workshop on Energy-Aware High Performance Computing · Frankfurt, GermanyHeld with the 2016 International Supercomputing Conference (ISC)
• The Jevons Paradox in Computing Systems ResearchDistinguished Lecture Series, Department of Computer Science, Virginia Tech
• Computational Significance and its Implications for HPC13th Workshop on Clusters, Clouds, and Data for Scientific Computing (CCDSC 2016) · Chemin de Chanzé, France
• Computational Significance and its Implications for Computing SystemsSchool of Electrical and Electronic Engineering, Newcastle University
• Scaling Up, Out, or DownSchool of Informatics, University of Edinburgh
• Significance-Driven Runtime SystemsRoMoL 2016 Workshop · Barcelona, Spain
• Advances in Energy-Efficient and Resilient HPC: Scaling Up, Out, or Back?Cardiff University

2015

• Why Energy-Efficient High Performance Computing is Harder than Energy-Efficient Embedded ComputingKeynoteWorkshop on Power & Energy-Aware High Performance Computing on Emerging Technology · Frankfurt, GermanyHeld with the 2015 International Supercomputing Conference (ISC)
• Variability: Why should we care?Birds-of-a-Feather session on Variability in Large-Scale Computing Systems (SC 2015) · Austin, TX
• New Approaches to Energy-Efficient and Resilient HPCDepartment of Computer Science, Old Dominion University
• HPDC Research at Queen’s: An OverviewARM High-Performance Computing Group · Manchester, U.K.
• Server Resource Provisioning for Real-Time Analytics using Iso-MetricsWorkshop on Performance Modeling: Methods and Applications (ISC 2015) · Frankfurt, Germany
• Evaluating Servers using Iso-Metrics: Power, Performance and Programmability Implications8th Workshop on Programmability Issues for Heterogeneous Multicores (MULTIPROG 2015) · Amsterdam, The Netherlands

2014

• The Challenges and Opportunities of Micro-Servers in the HPC Ecosystem12th Workshop on Clusters, Clouds, and Data for Scientific Computing (CCDSC 2014) · Chemin de Chanzé, France
• NVRAM as a User-Level Object StoreHiPEAC Autumn Computing Systems Week · Athens, Greece
• On the Viability of Microservers for Real-Time Data AnalyticsHiPEAC Autumn Computing Systems Week · Athens, Greece
• NanoStreams: A Hardware and Software Stack for Real-Time Analytics on Fast Data StreamsHorizon 2020: the HPC Opportunity · London, U.K.
• GEMSCLAIM: Greener Mobile Systems by Cross-Layer Energy ManagementCHIST-ERA Projects Seminar · Istanbul, Turkey

2013

• Programming the Energy Efficiency of HPC SystemsKeynote4th International Conference on Energy-Aware High Performance Computing · Dresden, Germany
• Connecting the Dots between Parallel Programming and EnergyKeynote21st Euromicro International Conference on Parallel, Distributed and Network-Based Processing, Belfast · Northern Ireland, UK
• Searching for Data: The Ever-Increasing Role of Memory Hierarchies on the Performance and Sustainability of Computing SystemsInaugural lecture, Queen’s University Belfast

2012

• Energy as a Resource in Parallel ProgramsSupercomputing 2012 Birds-of-a-Feather: Cool Supercomputing
• Block-Level Dynamic Dependence Analysis for Task-Based ParallelismWorkshop on Perspectives on Parallel Numerical Linear Algebra · Manchester, U.K.
• Software Techniques for Energy Conservation in High-End Computing SystemsInvited seminar, School of Computer Science, University of Manchester · U.K.

2011

• To Program or not to Program the Memory Hierarchy?KeynoteFourth Workshop on Programmability Issues for Heterogeneous Multicores (MULTIPROG) · Heraklion, Greece
• Energy Efficiency at Extreme Scale: Tools and ChallengesSupercomputing 2011 Birds-of-a-Feather: Energy-Efficiency

2010

• Rearchitecting MapReduce for Heterogeneous Multicore Processors with Explicitly Managed MemoriesSchool of Electronics, Electrical Engineering and Computer Science, Queen’s University Belfast

2009

• Determinism in Parallel Software and ArchitecturesHiPEAC Systems Week Cluster Meetings · Barcelona
• Parallelizing Non-trivial Applications with Multiple Programming ModelsHiPEAC Systems Week Cluster Meetings · Barcelona

2008

• Uniform Evaluation of Programming ModelsHiPEAC Systems Week Cluster Meetings · Paris

2007

• Unifying Layered Parallelism on the Cell BESupercomputing 2007 Birds-of-a-Feather: Unleashing the Power of the Cell Broadband Engine Processor for HPC
• Unified Scheduling of Polymorphic Parallelism on Asymmetric Multi-core SystemsLawrence Livermore National Laboratory
• System Software for Scaling on Many CoresOak Ridge National Laboratory

2006

• Facing the Challenges of Multicore Processor Technologies using Autonomic System SoftwareKeynote7th Workshop on Parallel and Distributed Scientific and Engineering Computing with Applications (PDSEC) · Rhodes, GreeceHeld with the 20th IEEE International Parallel and Distributed Processing Symposium (IPDPS), 1pp.
• Design and Implementation of Time-and Power-Efficient Software Stacks for Multicore ProcessorsIBM Thomas J. Watson Research Center
• Design and Implementation of Time-and Power-Efficient Software Stacks for Multicore ProcessorsDepartment of Computer Science, North Carolina State University
• Hardware Event-Driven Scalability Predictors: Improving Energy-Efficiency under Hard Performance Constraints on Multi-core and Multi-threaded ArchitecturesDepartment of Electronic and Computer Engineering, Technical University of Crete
• Addressing the Challenges of Chip Multiprocessors using Autonomic SoftwareDepartment of Computer Science, University of California, Riverside
• Addressing the Challenges of Chip Multiprocessors using Autonomic SoftwareDepartment of Electrical and Computer Engineering, University of British Columbia
• High-Performance Power-Efficient Runtime Environments for Dense Computing SystemsDepartment of Computer Science, Virginia Tech

2005

• High-Performance Power-Efficient Runtime Environments for Dense Computing SystemsInstitute of Computer Science, Foundation for Research and Technology — Hellas (FORTH)
• High-Performance Power-Efficient Runtime Environments for Dense Computing SystemsDepartment of Computer Engineering and Informatics, University of Patras

2004

• A Unified Programming Framework for Multigrain Multithreaded ArchitecturesInstitute of Computer Science, Foundation for Research and Technology — Hellas (FORTH)
• A Unified Programming Framework for Multigrain MultithreadingSchool of Electrical and Computer Engineering, National Technical University of Athens
• A Unified Programming Framework for Multigrain MultithreadingDepartment of Computer Science, University of California, Riverside
• A Unified Programming Framework for Multigrain Parallel ArchitecturesDepartment of Electrical and Computer Engineering, Northwestern University

2003

• Program Transformations and Scheduling Algorithms for Managing Shared Caches on Multithreaded ProcessorsDepartment of Informatics, Athens University of Economics and Business
• Program Transformations and Scheduling Algorithms for Managing Shared Caches on SMT ProcessorsIBM Thomas J. Watson Research Center
• Building Adaptive Programs with Local Sensing of Execution ConditionsDepartment of Computer Science, Texas A&M University

2002

• Interoperable System SoftwareDepartment of Information and Computer Sciences, University of California, Irvine
• Interoperable System SoftwareDepartment of Computer Science, College of William & Mary

2001

• Scaling Shared-Memory Programming Models beyond Shared-Memory ArchitecturesDepartment of Computer Science, University of Houston
• Some Steps towards Simple, Scalable and Portable Parallel Programming ModelsDepartment of Computer Science, College of William & Mary
• A Case for User-Level Page MigrationCoordinated Sciences Laboratory, University of Illinois at Urbana–Champaign