The Tesla K80 accelerator is a dual GPU board designed to cope with the most demanding computational workloads. It has 24GB of GPU accelerator memory meaning that the Tesla K80 delivers a 10x speed-up compared to the latest CPUs and TWICE the performance of its predecessor, the Tesla K20.
The K80 is based on the NVIDIA Kepler™ Architecture and the widely used CUDA® parallel computing model. The Tesla K-series GPUs are the leading platform for speeding through big data analytics and scientific computing workloads, which has a widespread ecosystem of software developers, software vendors and datacentre OEMs to accelerate findings and insight.