The symmetric multiprocessing (SMP)-based Cray CX1000-S configuration provides users with up to 128 cores of "big memory" computing built on Intel's QuickPath Interconnect (QPI) technology. The Cray CX1000-S system, which is one of the most dense large SMP solutions in the market, delivers maximum performance on scale-up jobs.
"Engineers, researchers and scientists want an affordable, mid-range high performance computing system that is architected with the latest technology, and the Cray CX1000-S system is designed to meet these evolving needs", stated Barry Bolding, vice president of Cray's products division. "With advanced CPU, GPU and now symmetric multiprocessing solutions based on today's launch of the latest Intel Xeon processors, Cray CX1000 customers can apply a customized supercomputing resource built with cutting-edge processing technology to their unique set of scientific challenges."
The Cray CX1000 supercomputer is priced starting at under $100,000, and provides high performance computing (HPC) users with a hybrid supercomputing architecture. Each configuration allows Cray CX1000 customers to architect a supercomputing system designed to solve computationally intensive problems with scale-out, scale-up or accelerated computing solutions.
"There is growing user demand for more affordable high performance computing", stated Raj Hazra, general manager of High Performance Computing at Intel. "Cray and Intel are providing opportunities for a whole new segment of the market to address complex problems with the Cray CX1000 based on the Intel Xeon processor E7 family."
Available in one to four chassis in a single, standalone cabinet, the three compute configurations for the Cray CX1000 system include the Cray CX1000-C, Cray CX1000-G and the Cray CX1000-S. The compute-based Cray CX1000-C features the dual-socket Intel Xeon Processor 5600 series for scale-out cluster computing; the Cray CX1000-G utilizes the NVIDIA Tesla GPUs for accelerator-based HPC; and the Cray CX1000-S features the Intel Xeon Processor E7 family. These complimentary technologies can be configured together to provide one system to meet a broad range of HPC application workloads.