This repo contains an implementation of finite difference method for heat equation with CUDA(Compute Unified Device Architecture).
- GPU : NVIDIA GeForce GTX 650 @ 1.072GHZ GDDR5 1GB
- CUDA toolkit 7.5
- CPU : Intel(R) Core(TM) i5-6400 @ 2.7GHZ
- RAM : DDR3L 16GB PC3-12800
- Microsoft Visual Studio Community 2013
TESLA C2070
: faster than a CPU approximately 60 ~ 90 timesNVIDIA GeForce GTX 650
: faster than a CPU approximately 10~20 times
- Current version is not optimized such as
shared memory
. - If you're interested in my works, please email me.