BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTAMP:20260623T031750Z
UID:f5d0cff9-e94b-4b6b-8d54-efb3c21cd101
DTSTART:20130909T083000Z
DTEND:20130911T153000Z
DESCRIPTION:Graphics Processing Units (GPUs) were originally designed to di
 splay\ncomputer graphics\, but they have developed into extremely powerful
 \nchips capable of handling demanding\, general-purpose calculations. The\
 nGPU architecture is inherently is more suited to many types of\nintensive
  parallel computations than the traditional CPU\, and hence\ncomputational
 ly demanding sections of code can be accelerated to\nsignificantly increas
 e overall performance. This is true not just for\nsmall-scale applications
  run on desktop size machines\, but also for\nthe largest-scale applicatio
 ns on massively parallel\narchitectures.\n\nApplications must be adapted t
 o utilise GPUs: most lines of\napplication source code are executed on the
  CPU and key computational\nkernels are distributed to the GPU cores. Curr
 ently\, for NVIDIA GPUs\,\nthe most popular programming method is the CUDA
  API\, which is\nextremely powerful but requires significant development e
 ffort. OpenCL\nis an alternative API\, which is less mature than CUDA but 
 has\nportability advantages. Recently\, a new higher-level standard has\ne
 merged\, OpenACC\, which promises to offer higher productivity. The\nprogr
 ammer uses "directives" in the code to provide the compiler with\nthe info
 rmation required to automatically offload code to the GPU. \n\nIn this 3-d
 ay course we will introduce and provide hands-on experience of\nCUDA\, Ope
 nCL (with more emphasis on the former) and OpenACC. In many\ncases it is r
 elatively straightforward to port a code to the GPU\, but\nmuch harder to 
 obtain good performance: we will cover a range of\ncommon GPU optimisation
  techniques.\n\nNo prior HPC or parallel programming knowledge is assumed\
 , but\nattendees must already be able to program in C\, C++ or Fortran. Ac
 cess\nwill be given to appropriate hardware for all the exercises.\n\nPre-
 requisite Programming Languages:\n\nFortran\, C or C++.\n\nhttps://events.
 prace-ri.eu/event/185/
SUMMARY:GPU Programming with CUDA and OpenACC @ EPCC
URL;VALUE=URI:https://events.prace-ri.eu/event/185/
END:VEVENT
END:VCALENDAR