Home Products GPU Page for continuity

GPU Computing - A paradigm shift

nVidia Tesla GPU Workstation

Need Powerful Server Options?

Are you looking to deliver equivalent performance at 1/10th the cost and 1/20th of the power consumption? Enter the arena of GPU computing.

GP-GPU computing has become the latest disruptive force to take HPC by storm. By leaving sequential computations to the CPU, and moving parallel computations to the GPU; the potential for enhanced performance increases exponentially. Some applications are upwards of 250 times faster than comparable performance on a CPU processor alone.

From facilitating your proof of concept/research for your applications compatibility with the GPU processor to building a cutting-edge heterogeneous supercomputer - Red Barn has you covered.

GPU Heterogeneous Supercomputing

Features

  • Support for the entire CUDA-enabled NVIDIA and ATI FireStream product lines
  • Industry standard, non-proprietary hardware
  • NVIDIA Cuda Drivers and Development kit pre-installed
  • Fully-customizable personal super computers, servers and turn-key clusters to best meet the needs of your applications
  • Up to 4GPUs in a workstation, and 8GPU in a server
  • Custom exterior logos
Back to Top

Tesla GPU Computing Products

RBC Heterogeneous Clusters

  • Turn-key cluster solutions designed to meet your needs
  • Pre-sales consultation to assist you in maximizing your applications
  • Customized chassis designs available to meet specific CPU/GPU ratios
  • Tested rackmount configurations to optimize bandwidth on Intel and AMD platforms

RBC GPU Servers

  • Many form factors available -1U to 4U
  • Up to 4 GPU in 1U form factor
  • Up to 8 GPU in 4U form factor
  • Built on Intel or AMD platform
  • Can integrate all CUDA enabled GPUs into the servers (Tesla, GeForce or Quadro)
  • Options available to attach Nvidia-enabled All-Tesla computing systems to host servers to accommodate specific CPU/GPU ratios.

RBC-HPC Personal Supercomputer

  • Fully customizable configurations
  • Intel Core i7, Xeon or AMD Opteron platforms
  • Up to 4 double-width GPU cards of any variety accepted
  • Up to 192GB memory
  • Dual Intel-chipset
  • Up to 7 PCI-E expansion slots
  • Ultra-Quiet chassis available for ofice environments
nVidia Tesla GPU Accelerators Back to Top

Founded in 1996, Red Barn Computers has become a leader in High Performance Computing and Cluster Applications. Our expertise has given us the unique ability to offer total solutions encompassing all aspects of HPC - including turn-key solutions, storage, design, and administration. Red Barn also offers a wide array of custom Linux or open architecture platforms that are designed and built using high-end, reliable, industry standard components. Our products include Dual, Quad, and 8 CPU Systems, along with high end workstations. Our Clustering products are based on the Warewulf/Perceus platform, and provide numerous solutions for High Performance and High Availability Clusters.