BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000713Z
LOCATION:702
DTSTART;TZID=America/Denver:20231113T103000
DTEND;TZID=America/Denver:20231113T105400
UID:submissions.supercomputing.org_SC23_sess449_ws_pawatm102@linklings.com
SUMMARY:Implementing Scalable Matrix-Vector Products for the Exact Diagona
 lization Methods in Quantum Many-Body Physics
DESCRIPTION:Tom Westerhout (Radboud University) and Bradford L. Chamberlai
 n (Hewlett Packard Enterprise (HPE))\n\nExact diagonalization is a well-es
 tablished method for simulating small quantum systems. Its applicability i
 s limited by the exponential growth of the Hamiltonian matrix that needs t
 o be diagonalized. Physical symmetries are usually utilized to reduce the 
 matrix dimension, and distributed-memory parallelism is employed to explor
 e larger systems. This paper focuses on an implementation of the core dist
 ributed algorithms, with a special emphasis on the matrix-vector product. 
 Instead of the conventional MPI+X paradigm, Chapel is chosen as the langua
 ge in this work.\n\nWe provide a comprehensive description of the algorith
 ms and present performance and scalability tests. Our implementation outpe
 rforms the state-of-the-art MPI-based solution by a factor of 7--8 on 32 c
 ompute nodes or 4096 cores and scales well through 256 nodes or 32768 core
 s. The implementation has 3 times fewer software lines of code than the cu
 rrent state of the art, but is still able to handle generic Hamiltonians.\
 n\nTag: Accelerators, Artificial Intelligence/Machine Learning, Algorithms
 , Applications, Architecture and Networks, Distributed Computing, Compiler
 s, Data Analysis, Visualization, and Storage, Exascale, Heterogeneous Comp
 uting, Linear Algebra, Message Passing, Performance Optimization, Programm
 ing Frameworks and System Software, Quantum Computing, Runtime Systems, So
 ftware Engineering, Sustainability, Task Parallelism, Tensors\n\nRegistrat
 ion Category: Workshop Reg Pass\n\nSession Chairs: Engin Kayraklioglu (Hew
 lett Packard Enterprise (HPE)), Daniele Lezzi (Barcelona Supercomputing Ce
 nter (BSC)), Bill Long (Retired), Karla Vanessa Morris Wright (Sandia Nati
 onal Laboratories), Irene Moulitsas (Cranfield University), and Elliott Sl
 aughter (SLAC National Accelerator Laboratory)\n\n
END:VEVENT
END:VCALENDAR