BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20241120T082410Z
LOCATION:HG F 30 Audi Max
DTSTART;TZID=Europe/Stockholm:20240604T095800
DTEND;TZID=Europe/Stockholm:20240604T095800
UID:submissions.pasc-conference.org_PASC24_sess158_pos161@linklings.com
SUMMARY:P44 / ACMP03 - Scalable Simulations of Resistive Memory Devices: A
  Dynamical Monte Carlo Approach
DESCRIPTION:Poster\n\nAlexander Maeder, Manasa Kaniselvan, Marko Mladenovi
 ć, Mathieu Luisier, and Alexandros Nikolaos Ziogas (ETH Zurich)\n\nResisti
 ve random access memories (ReRAM) are expected to play a prominent role in
  modern computer architectures due to their low cost, simple structure, an
 d unique functionality. The long-range atomic movements inside these devic
 es, which occur over extended timescales under applied fields, can be accu
 rately described by Dynamical Monte Carlo (DMC) simulations. In DMC, the c
 ontinuum movements of atoms are discretized into ‘events’ on an atomistic 
 graph, which is time-stepped under the influence of external fields (poten
 tial, Joule heating). Parallelization can only occur within each step, ren
 dering such simulations highly sensitive to data movement. Here, we presen
 t a scalable DMC code that simultaneously optimizes the different computat
 ional kernels found in the field solvers (systems of linear equations, mat
 rix-vector multiplication) and event selection (prefix sums). Our implemen
 tation leverages preconditioned sparse iterative solvers, graph-based doma
 in decomposition to divide work between nodes, and hybrid CPU-GPU computat
 ions to optimize node usage and data transfer in a distributed environment
 . The acceleration ultimately enables the first investigation of ReRAM cro
 ssbar arrays with an atomistic resolution, providing deeper insights into 
 the operating mechanisms of these devices and paving the way for their mai
 nstream adaptation in future memory technologies.\n\nSession Chair: Iva Ka
 vcic (Met Office)
END:VEVENT
END:VCALENDAR
