HPC Solutions

High Performance SSD Cluster and GPU Cluster Solutions

High Performance Computing is experiencing a radical transformation as it evolves from dedicated server clusters to hybrid GPU computing. NextIO is at the forefront of this transformation with industry leading vCORE GPU compute systems, and vSTOR external SSD enclosure appliances.

Researchers are constantly pushing the hardware performance envelope with custom developed applications for seismic modeling, predictive analysis for oil and gas exploration, bioinformatics and other compute-intensive jobs. GPU cluster compute, standalone GPU compute and hybrid compute are mainstream hardware solutions designed to cluster as many CPU cores as possible for solving scientific problems.

Lowering I/O Latency with NextIO HPC GPU and Flash Storage Technology

Low latency for these systems has traditionally focused on the network interconnecting the cluster nodes. However, the data sets themselves are typically very large and they must be broken up and run over several systems in parallel. With up to 20TB of storage and potentially over 4 million IOPS throughput, vSTOR enables scientists to fully leverage their CPU and GPU investments by providing as much high-speed, low-latency data as the cores can process. NextIO HPC GPU solutions support from 1 to 16 GPU's per server enabling as much GPU compute power as the application demands in 1U and 3U form factor.

To better understand why GPU compute systems are substituting dedicated supercomputers, download the case study, "NextIO vCORE™ Express 2070 Dramatically Accelerates Space & Upper Atmosphere Simulations".