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:20191206T142452Z
LOCATION:Analog 1\, 2
DTSTART;TZID=Europe/Stockholm:20180625T132500
DTEND;TZID=Europe/Stockholm:20180625T132900
UID:isc_hpc_ISC High Performance 2018_sess103_phd117@linklings.com
SUMMARY:(PhD07) pystencils - Automatic Generation, Optimization and Analys
 is of Stencil Codes
DESCRIPTION:PhD Forum\n\n(PhD07) pystencils - Automatic Generation, Optimi
 zation and Analysis of Stencil Codes\n\nHönig, Bauer\n\nWe present a Pytho
 n-based metaprogramming toolkit for stencil codes called pystencils.\npyst
 encils reduces development and maintenance costs, increases readability an
 d reusability of compute-intensive numerical kernels by introducing high l
 evel abstractions.  \n\nThis is achieved with a domain-specific language (
 DSL) based on SymPy.\nA DSL in SymPy provides not only an extensible and f
 lexible model formulation, but also a computer algebra system, which is ve
 ry useful for the manipulation of the problem description. The model can b
 e described on multiple abstraction levels, either directly as a stencil c
 ode, or preferably as a pure mathematical formulation which is then automa
 tically discretized and transformed into a stencil description. \nThis ste
 ncil description is represented as an abstract syntax tree (AST), which un
 dergoes multiple transformations. The toolkit provides optimizing transfor
 mations, that operate on the AST, like loop splitting/merging, SIMD vector
 ization or common subexpression elimination.\nThe AST is then passed to on
 e of the backends to generate C++, CUDA or LLVM code. \nFor performance ev
 aluation, the generated kernel can be automatically benchmarked and examin
 ed in detail with the help of kerncraft - a loop kernel analysis and perfo
 rmance modeling toolkit. Generated kernels can be integrated into existing
  C/C++ frameworks like waLBerla, or run directly from Python. Direct execu
 tion in Python allows fast prototyping, since all of Python's powerful uti
 lities and libraries are available.\n\nAll these features make pystencils 
 a powerful tool for systematic performance engineering of numerical comput
 e kernels that simplifies or automates important steps in development, ana
 lysis and optimization of the kernel.\n\nPasses: Conference Pass, Exhibiti
 on Pass, Compilers, HPC workflows, Performance Analysis and Optimization, 
 Scientific Software Development\n\nTag: Conference Pass, Exhibition Pass, 
 Compilers, HPC workflows, Performance Analysis and Optimization, Scientifi
 c Software Development
URL:https://2018.isc-program.com/presentation/?id=phd117&sess=sess103
END:VEVENT
END:VCALENDAR