GPU610/DPS915 Glossary

From CDOT Wiki
Revision as of 18:52, 6 January 2017 by Chris Szalwinski (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

GPU610/DPS915 | Student List | Group and Project Index | Student Resources | Glossary

Please add your term to the end of the table using the following syntax!


| TERM || DEF || [[User:WUN | FN]]|| LN || CREDIT |-
  • TERM: the term being defined
  • DEF: definition of the term
  • WUN: your Wiki user name
  • FN: your first name
  • LN: your last name
  • CREDIT: the original credit (if any)

List of Terms

Term Definition Enetered by Last Name Credit
Host The serial processing part of a heterogeneous computer Chris Szalwinski -
Device The parallel processing part of a heterogeneous computer Chris Szalwinski -
ulp Unit in last place - unit of least precision - spacing between adjacent floating point numbers Chris Szalwinski Wikipedia July 15 2012
Heterogeneous computer Computer that has more than one type of processor (eg. CPU & GPU) James Boelen Chris Szalwinski
NVIDIA Producer of GPUs that contain CUDA cores James Boelen Chris Szalwinski
HSA Heterogeneous system architecture foundation - collection of hardware manufacturers creating heterogeneous computers James Boelen Chris Szalwinski
STI Sony Toshiba IBM - Alliance to create and develop the cell processor (combination CPU GPU) James Boelen Chris Szalwinski
Open MPI Open Message Passing Interface James Boelen Chris Szalwinski
Open MP Open Multi-Processing James Boelen Chris Szalwinski
Cilk Plus extension to c/c++ to be able to program heterogeneous computers (championed by intel) James Boelen Chris Szalwinski
CUDA CUDA™ is a parallel computing platform and programming model that enables dramatic increases in computing performance by harnessing the power of the graphics processing unit (GPU) Bhrugesh Patel NVIDIA
OpenCL Platform independent standard, developed by the Khronos group. Set of extensions you can add to c/c++ to do parallel programming. (complicated and hard to learn) James Boelen Chris Szalwinski
OpenACC #pragmas that leaves parallelization to the compiler James Boelen Chris Szalwinski
DirectCompute Microsoft’s version of OpenCL / CUDA James Boelen Chris Szalwinski
C++ AMP MS compiler directives (similar to OpenACC) James Boelen Chris Szalwinski
Multi processors processor technology that supports concurrency-oriented programming James Boelen Chris Szalwinski