SC26 Committee Awarded Test of Time Honor
Nathan Bell and Michael Garland were honored for their influential research on sparse matrix multiplication.
Updated on Sept. 29, 2026 in Mathematics
Live Poll
Should scientific research be evaluated primarily by its long-term influence rather than immediate project results?
The SC26 committee selected the 2009 research paper "Implementing Sparse Matrix-Vector Multiplication on Throughput-Oriented Processors" for the prestigious Test of Time Award. The honor recognizes the work's enduring impact on high-performance computing over the past 17 years.
Why it matters
The award highlights research presented at the SC conference series that continues to shape technical fields long after its initial publication. This study specifically pioneered techniques for managing irregular data structures on graphics processors, which remain central to modern throughput-oriented computing.
The researchers utilized an NVIDIA GeForce GTX 285 to evaluate sparse matrix formats including DIA, ELL, CSR, and COO. Their methodology focused on optimizing computation to mitigate execution and memory divergence caused by irregular matrix structures.
The players
Nathan Bell
He is a Principal Engineer at Google who co-authored the award-winning research paper.
Michael Garland
He is the Senior Director of Programming Research at NVIDIA and a long-time contributor to GPU computing research.
SC26
This is the international supercomputing conference where the committee formally announced the award recipients.
The details
The study organized computational tasks to overcome challenges posed by the irregular data structures inherent in sparse matrices. By refining these processes for throughput-oriented processors, the authors established foundational techniques for parallel computing efficiency.
Timeline
Michael Garland joined NVIDIA in 2006.
The original paper was presented at the SC09 conference in 2009.
The Test of Time Award was officially announced on September 29, 2026.
The Big Picture
The recognition of this 2009 paper validates the transition toward throughput-oriented architectures in high-performance computing. It confirms the long-term importance of the SC09 conference as a catalyst for GPU-based computational breakthroughs.
The methods detailed in this research provide the structural foundation for modern GPU-accelerated computing across diverse scientific fields. These optimizations enable faster processing speeds for complex mathematical models in consumer and research technologies.
The takeaway
This award underscores the persistent value of foundational research in optimizing hardware for complex data tasks. Readers can take away that high-performance breakthroughs often require long-term commitments to refining basic computational efficiency.
What happens next
Michael Garland is scheduled to present a talk titled Twenty Years of Sparsity on GPUs at the upcoming SC26 conference.
Further reading
Explore more advancements in Mathematics on our dedicated research page.
More information
View the Original research paper access to understand the technical methodology.
Source note: This article includes information reported by HPCwire.
Live Poll
Should scientific research be evaluated primarily by its long-term influence rather than immediate project results?
