In a cloud environment, GPUs work in conjunction with a server’s CPU to accelerate application and processing performance. The CPU offloads compute-intensive portions of the application to the GPU which processes large blocks of data at one time rather than sequentially boosting the overall performance in a server environment.
GPUs are better for high performance computing than CPU’s alone because of the thousands of small efficient cores designed to process information faster.
JPStream customers can choose the NVIDIA Tesla or GRID graphic cards that best meet the needs of their workloads. NVIDIA Tesla is designed for high performance acceleration of scientific computation and data analytics, and NVIDIA GRID is engineered for professional grade virtualized graphics.