GPU-as-a-Service

On May 9, ScaleMatrix announced a GPU-as-a-Service solution with the dense Graphics Processing Unit (GPU) power to deliver unprecedented efficiency gains at an affordable price point for today’s demanding HPC, Deep Learning and Artificial Intelligence (AI) workloads.
 

This solution provides organizations a better, faster and less expensive alternative to Amazon P2 Instance, at a competitive price/performance point. With GPU-as-a-Service, ScaleMatrix supplies the servers as a bare-metal cloud offering with flexible payment options and short-term use models for GPUs on premium servers.  The solution delivers dedicated direct access with the compliance and security users need without the cloud or virtualization overhead. In addition, ScaleMatrix delivers up to a 30% improvement on Power Usage Effectiveness (PUE).
 

The system specifications range from 4-8 GPUs, with up to dual Intel® Xeon® processor E5-2630 v4, between 64-128GB RAM and a 1TB SSD NVMe for the primary drive, and 4TB SATA as a secondary drive.  The GPU-as-a-Service solutions start as low as $1,464 per month and scale to meet the needs of the consumer’s growing demand for intense GPU compute processing.

ScaleMatrix also offers technical expertise in the areas of HPC, Deep/Machine Learning and AI with our Matrix TotalCare™ service that provides seasoned data center experts who take the guesswork out of remote hands support and ensure timely responses to all technical inquiries.

If you happen to be at the GPU Technology Conference (GTC 2017) in San Jose this week, visit us in Booth #528 to learn more.