2026 SIGHPC Doctoral Dissertation Award Winner 

2026 SIGHPC Doctoral Dissertation Award Winner  

ACM's Special Interest Group on High Performance Computing (SIGHPC) is pleased to announce that Dr. J. Gregory Pauloski has won the 2026 SIGHPC Doctoral Dissertation Award. This award is given each year for the best doctoral dissertation completed in high-performance computing (HPC) in the previous year. Nominations were evaluated on the novelty of the work, quality of scholarship, significance of the research contributions, and potential impact on theory and practice. The award includes a $2,000 cash prize, a plaque, and recognition at the International Conference for High Performance Computing, Networking, Storage, and Analysis (SC'26) in November 2026.

Dr. Pauloski's dissertation confronts the challenge of programming distributed science applications across the "computing continuum" — the increasingly heterogeneous mix of personal, cloud, edge, and high-performance computing systems on which modern scientific workflows run. His dissertation centers on ProxyStore, a library that extends the pass-by-reference programming model to distributed applications, decoupling data flow from control flow so that the method, destination, and timing of data movement can be selected dynamically — a long-standing challenge in distributed computing. To rigorously evaluate this approach, he also developed TaPS, a benchmarking suite for task-based parallel execution frameworks. Building on the object-proxy paradigm, he introduced three higher-level patterns — distributed futures, streaming, and ownership — that give developers reusable tools for recurring needs in scientific workflows ranging from bioinformatics to materials science: overlapping computation and communication, decoupling event notification from bulk data transfer, and safely managing memory across distributed processes. Most recently, his work on Academy, a federated agents system, supports the creation and deployment of pervasive autonomous agents that decentralize control flow across stateful entities, extending these ideas toward agentic scientific workflows. Together, this body of work delivers an open-source toolbox and set of frameworks for developing performant science applications across the computing continuum.

“AI is changing how scientists build applications, and that means rethinking how those applications interact with HPC,” said Dr. Pauloski. "Increasingly, they are dynamic, adaptive, and more reliant on federated resources rather than any single system. From ProxyStore to Academy, my research has focused on coordinating data and control flow across these heterogeneous resources, so scientists can build applications that scale seamlessly from laptops to supercomputers. I hope this work lays the groundwork for agentic systems that can leverage all the resources scientists have access to, in performant and reliable ways."

Dr. Pauloski received his Ph.D. in Computer Science from the University of Chicago, where he was co-advised by Ian Foster and Kyle Chard of Globus Labs. He is currently a Senior Engineer at NVIDIA, where he works on AI/ML infrastructure.

The ACM SIGHPC Doctoral Dissertation Award committee was chaired by Shantenu Jha (Rutgers-New Brunswick and Princeton Plasma Physics Lab) and composed of Douglas Thain (University of Notre Dame), Jeffrey A. F. Hittinger (Lawrence Livermore National Laboratory), Ana Gainaru (Oak Ridge National Laboratory), Ana-Lucia Varbanescu (University of Twente), Dejan Milojicic (Hewlett Packard Labs), Allen D. Malony (University of Oregon), and Rio Yokota (Institute of Science Tokyo).