This latest-generation Nvidia graphics card is based on Pascal architecture. It can also be used as a graphic accelerator for 3D software, or even as a co-processor, thanks to CUDA cores that perform single-precision processing, such as Deep-Learning or video encoding.
GeForce GTX 1080 Ti:
- 3584 CUDA cores
- 11 GB GDDR5X
- 484 GB/s maximum bandwidth
- Nvidia GPU Boost 3.0 Technology
Why GPU?
A GPU (Graphical Process Unit), has a massively parallel architecture compared to a conventional CPU. This design originally intended to handle computation only for computer graphics (OpenGL/Direct3D), was adapted with programming languages such as CUDA or OpenCL, to perform computation in applications traditionally handled by CPU.
Due to its ability to speed up processing, GPU architecture is particularly suited for resource-intensive applications which require a large number of parallel tasks (image processing, bioinformatics, big data … ).