BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Sparse Linear Algebra as a Computational Paradigm for Irregular Sc
 ientific Computation
DTSTART:20261014T141500
DTEND:20261014T151500
DTSTAMP:20261002T231645Z
UID:48db9f08b4b1a2defa10b72ccc9f2e520b8aa789c72089e8b5bc335b
CATEGORIES:Conferences - Seminars
DESCRIPTION:Giulia Guidi\, Cornell University\nFrom life sciences to data 
 analytics and machine learning\, a growing class of workloads involves irr
 egular\, data-intensive computation that maps poorly onto modern hardware\
 , yet increasingly requires access to high-performance computing resources
 . This talk presents sparse linear algebra as a unifying computational par
 adigm for such problems. By reformulating a computation in terms of a smal
 l set of primitives\, such as sparse matrix multiplication\, matrix-vector
  products\, and triangular solves\, these computations become tractable at
  scale and portable across architectures\, while hardware-specific tuning 
 is delegated to expert-maintained libraries. In this talk\, I will focus p
 rimarily on recent work on Mikado\, which recasts a core operation in biob
 ank-scale population genetics\, traditionally a traversal of a large direc
 ted acyclic graph\, as a sparse triangular solve. Under a topological orde
 ring\, the graph’s adjacency matrix is strictly lower triangular\, and t
 he traversal reduces to pipelined sparse matrix-vector products\, cutting 
 hours or days of CPU time down to minutes on a single GPU without custom k
 ernels. I will conclude with open problems in the automatic discovery of s
 parse structure in scientific codes\, its implications for hardware co-des
 ign\, and the role of sparse computation in AI for science.\n 
LOCATION:CM 1 517 https://plan.epfl.ch/?room==CM%201%20517
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR