NextIO has released the vCORE Express Lite GPU computing solution, designed with two Fermi-based Nvidia Tesla 20-Series computing processors with 12GB of total GPU RAM in a standard 1U chassis. Compared to multi-core CPUs, the vCORE Express Lite is reported to deliver the same performance at 10 per cent the cost and five per cent the power consumption. Designed for the technical and enterprise workgroup computing space, it delivers Error Correction Codes (ECC) memory for accuracy and scalability, and offers full GPU communication bandwidth.
With two GPUs, the cost of building a test or starter cluster is reduced while maintaining the flexibility to test different configurations on the server. For projects that cannot effectively utilise multiple GPUs per server, the two-GPU option offers the ability to provide only the capacity that is needed for a specific project. The vCORE Express Lite also allows the GPU and the server to be managed separately while minimising downtime, and is suited for user with a low number of job counts, low-to-medium GPU counts, or a combination of both