NextIO, the premier provider of next-generation I/O solutions, and NVIDIA, the inventor of the GPU, today announced a partnership designed to accelerate CUDA development and deployment using NextIO Adaptive Connect® PCIe chassis and management software. NextIO’s technology allows servers to take full advantage of two or more GPUs at PCIe speeds, enabling the application of multi-GPU solutions to high-performance compute problems.

The relationship provides customers with a broader set of solutions for developing multi-GPU applications on a single platform. The NextIO management framework allows servers to be mapped across PCIe up to 10 GPUs simultaneously. Server connections can be changed “on the fly” based on development or application demands.

NextIO will deliver pre-packaged GPU solutions to customers requiring a fully supported and robust CUDA development and deployment platform utilizing NVIDIA Tesla and Quadro GPU products.

“Our customers in the oil and gas, medical imaging, and simulation markets have been demanding low-cost, high-performance GPU solutions from NextIO as a means to scale their compute applications to 50 percent or greater workloads,” said K.C. Murphy, CEO, NextIO. “Our partnership with NVIDIA will provide turnkey supported NVIDIA solutions to those customers looking for scale-out solutions.”

“Our partnership with NextIO enables customers to realize better economies of scale in a pre-packaged GPU solution,” says Jeff Brown, general manager of professional solutions, NVIDIA. “Now professionals developing and deploying applications on NVIDIA Quadro and Tesla GPUs are no longer constrained to one or two GPU’s per server, allowing more extensible GPU solutions.”