vCORE Extreme: for HPC GPU and GPU cluster requirements

vCORE Extreme

vCORE Extreme: Increasing GPU compute density for GPU supercomputing

Scale-Out HPC GPU Computing Solution

NextIO vCORE™ Extreme provides high density consolidated GPU hybrid computing capabilities by presenting configurable pools of manageable GPU resources to low-cost servers.

vCORE Extreme is a powerful HPC GPU solution for consolidated GPU computing systems and enabling dynamic, large scale HPC computing configurations. vCORE Extreme is scalable, flexible, and serviceable.

  • Ability to add GPU's on demand
  • Serviceable GPUs, redundant power and cooling
  • Management software enables dynamic GPU reconfigurability
  • Dense GPU platform delivers increased GPU supercomputing performance

Increase Your GPU Compute Density for GPU Supercomputing

NextIO vCORE™ Extreme enables you to maximize your rack and server density by moving GPU's out of servers and into a 3u vCORE Extreme GPU chassis. Each vCORE Extreme supports up to 16 double-wide GPU's and 8 server connections over x16 PCIe links enabling scale-out or scale-up GPU computing delivered on-demand.

Now each 1u server can access from 1 to 16 GPU's without burdening the server with the cost overhead of GPU's, extra power and cooling. When GPU's are moved out of servers to vCORE Extreme, higher server to GPU densities are realized, performance is increased, and costs are reduced.

Get Faster Results

Most of today's 1u servers support a maximum of two internal GPU's. By expanding the number of GPU's available to a job, performance is increased and job time is reduced. Powered by NVIDIA Tesla, vCORE Extreme provides the highest level of supercomputing capability right out of the box.

Reduce Your Job Costs

Running more jobs in less time not only increases productivity, but can also increase your revenue. As significant as this is it's not the whole story. vCORE Extreme provides a centralized, shared pool of GPU resources allowing a quick and easy way to manage your GPU's and further lower your total cost of ownership. Pools of GPU resources can be quickly reconfigured among servers in the rack and each job can be optimized for the right ratio of CPU's to GPU's without over provisioning servers.

vCORE Extreme is purpose-built for hosting GPU's. Each GPU is housed in a hot-plug tray for ease of service and upgrades with minimum downtime. Utilizing 92mm hot-plug fans and four 1400W high-efficiency hot-plug PSU's, vCORE Extreme uses less power and provides better cooling than alternative solutions.

I/O Management For Your GPU Pools

Ease of upgrades and serviceability keep your projects running even during an event. The dedicated vCORE Extreme Intelligent Management Platform Interface (IPMI) port provides immediate status and event notification allowing you to service the unit quickly and easily using hot-plug interfaces. vCORE Extreme supports third party cluster management applications for maximum GPU cluster control.