Mobirise

15th Workshop on Irregular Applications: Architectures and Algorithms

November 16, 2025
America’s Center, Saint Louis, MO
Room 232
In conjunction with SC25

Mobirise

Program

TimeEvent
9:00 - 9:05Welcome and Introduction
Giulia Guidi (Cornell University), Flavio Vella (Università di Trento), Antonino Tumeo (PNNL)
9:05 - 10:00Invited Talk 1
Chair: Antonino Tumeo (PNNL)

Maintaining connected components for unending graph streams
Cynthia A. Phillips (Sandia National Laboratories)
10:00 - 10:30
Coffee Break
10:30 - 12:20


10:30 - 10:55


10:55 - 11:20


11:20 - 11:45


11:45 - 12:10


12:10 - 12:20
Paper Session 1: GPU Graph and Parallel Algorithms
Chair: Vito Giovanni Castellana (PNNL)

eIM: GPU-Accelerated Efficient Influence Maximization in Large-Scale Social Networks
Doney, Huang, Lee

Profiling Application-Specific Properties of Irregular Graph Algorithms on GPUs
Sharma, Burtscher

Generating Permutations at Scale
Green, Eaton, Tripathy, Nolet, Luitjens

Performance-Portable Symbolic Factorization through Common Graph Operation
Selvitopi, Li, Buluc

How Effective Is Matrix Reordering for Improving Performance of Sparse Matrix-Vector Multiplication? [short]
Asudeh, Mahdipour Saravani, Rastello, Sabin, Sadayappan
12:20 - 14:00Lunch Break
14:00 - 15:00

Invited Talk 2
Chair: Giulia Guidi (Cornell University)

Communicating more—asynchronously—saves time
Rich Vuduc (Georgia Institute of Technology)
15:00 - 15:30Coffee Break
15:30 - 12:00


15:30 - 15:55


15:55 - 16:20


16:20 - 16:45


16:45-16:55
Paper Session 2: Architectures and Accelerators
Chair: Ankur Limaye (PNNL)

An Optimized Generalized Multi-Color Point Implicit Solver for Intel GPUs using OneAPI ESIMD
Wassell, Zubair, Walden, Nastac, Nielsen, Ewart

Architecting Tensor Core-Based Reductions for Irregular Molecular Docking Kernels
Solis-Vasquez, Tillack, Santos-Martins, Koch, Forli

Comparing Graph Algorithm Styles on NVIDIA and AMD GPUs
VanAusdal, Burtscher

Benchmarking and Dissecting the Cerebras Wafer Scale Engine-2 Architecture [short]
Miyajima, Fukuoka
16:55 - 17:25Debate: Static algorithms, dynamic graphs
Chair: Flavio Vella (Università di Trento)

Proposition: “Static graph algorithms are a solved problem — both algorithmically and architecturally.
The current systems and algorithmic approaches are inappropriate for dynamic networks.”

Panelists: Anastasiia Butko (LBNL), Sanjukta Bhowmick (University of North Texas),
Michela Taufer (University of Tennessee, Knoxville), Albert-Jan Yzelman (Huawei)

17:25 - 17:30Closing Remaks
Giulia Guidi (Cornell University), Flavio Vella (Università di Trento), Antonino Tumeo (PNNL)

HTML Maker