BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
 M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20191113T223807Z
DTSTART:20190920T130000Z
DTEND:20190920T140000Z
SUMMARY:Seminar: Parsl: Pervasive Parallel Programming in Python
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}bee-k05dgdm
 s-ugesoi
DESCRIPTION:Join us for the next Computer Science Atlas talk with Daniel 
 S Katz from the University of Illinois\, USA.\n\n\nHigh-level programmin
 g languages such as Python are increasingly used to provide intuitive in
 terfaces to libraries written in lower-level languages and for assemblin
 g applications from various components. This migration towards orchestra
 tion rather than implementation\, coupled with the growing need for para
 llel computing (e.g.\, due to big data and the end of Moore's law)\, nec
 essitates rethinking how parallelism is expressed in programs. Here\, we
  present Parsl\, a parallel scripting library that augments Python with 
 simple\, scalable\, and flexible constructs for encoding parallelism. Th
 ese constructs allow Parsl to construct a dynamic dependency graph of co
 mponents from a Python program enhanced with a small number of decorator
 s that define the components to be executed asynchronously and in parall
 el\, and then execute it efficiently on one or many processors. Parsl is
  designed for scalability\, with an extensible set of executors tailored
  to different use cases\, such as low-latency\, high-throughput\, or ext
 reme-scale execution. We show\, via experiments on the Blue Waters super
 computer\, that Parsl executors can allow Python scripts to execute comp
 onents with as little as 5 ms of overhead\, scale to more than 250000 wo
 rkers across more than 8000 nodes\, and process upward of 1200 tasks per
  second. Other Parsl features simplify the construction and execution of
  composite programs by supporting elastic provisioning and scaling of in
 frastructure\, fault-tolerant execution\, and integrated wide-area data 
 management. We show that these capabilities satisfy the needs of many-ta
 sk\, interactive\, online\, and machine learning applications in fields 
 such as biology\, cosmology\, and materials science.\n
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
LOCATION:Kilburn L.T 1.5\, Kilburn Building\, Manchester
END:VEVENT
END:VCALENDAR