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:20210916T132456Z
LOCATION:
DTSTART;TZID=Europe/Stockholm:20210706T173000
DTEND;TZID=Europe/Stockholm:20210706T190000
UID:submissions.pasc-conference.org_PASC21_sess182_post136@linklings.com
SUMMARY:P19 - Scalable Distributed Memory Implementation of Dissipative Qu
 antum Dynamics Subject to a Non-Markovian Environment
DESCRIPTION:Poster\n\nP19 - Scalable Distributed Memory Implementation of 
 Dissipative Quantum Dynamics Subject to a Non-Markovian Environment\n\nOvc
 harenko, Fingerhut\n\nThe dynamics of a quantum system in contact with env
 ironment is central to the understanding of numerous processes, e.g., the 
 ultrafast dynamics of photoexcited biological systems or operation of qubi
 ts in quantum computers. Established numerical methods for the real time d
 escription of dissipative quantum dynamics, like, the Hierarchical Equatio
 ns of Motion are suited for the weak-to-intermediate system-bath coupling 
 regime. In contrast, the QUasi-Adiabatic Path Integral (QUAPI) formalism h
 as its starting point in the strong coupling limit. Numerical challenges o
 f the QUAPI method are rooted in the exponentially growing number of paths
  during propagation. The recently introduced mask assisted coarse graining
  of the Feynman-Vernon influence coefficients (MACGIC-QUAPI) provides acce
 ss to biological relevant system sizes and long-time system-bath correlati
 ons.[1] Nevertheless, MACGIC-QUAPI simulations remain challenging. Especia
 lly for distributed memory systems, the necessity of fast lookup of specif
 ic paths distributed over computing nodes poses challenges. We present a s
 calable MPI-TBB parallel implementation of the MACGIC-QUAPI method. Exploi
 ting task-based parallelism and hash-table based path search provides effi
 cient use of computer resources and facilitates inter-node load balancing.
  Benchmark calculations are compared to the performance of boost-MPI and O
 penMP thread-based implementations. [1] M.Richter and B.Fingerhut, J.Chem.
 Phys. 146, 214101 (2017) and Faraday Discuss. 216, 72 (2019).
END:VEVENT
END:VCALENDAR