Skip to content
/ DFGPU Public

Investigating if double-precision values can be sped up on NVIDIA GPUs through the use of extended precision, and what the tradeoffs are.

Notifications You must be signed in to change notification settings

kw510/DFGPU

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DFGPU

This research project investigates the use of the software technique Extended Precision for the purpose of a near double precision floating point number. The performance and correctness were tested against the native double floating point number upon the GPU.

A research paper was then produced, which shows the results of the project.

About

Investigating if double-precision values can be sped up on NVIDIA GPUs through the use of extended precision, and what the tradeoffs are.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published