GPU Rack

In ultracompetitive industries and fields of research, organizations are under pressure to process large volumes of data and increasingly complex computational tasks and to provide their professionals with the hardware and tools they need to innovate and lead. Investing in compact, energy-efficient GPU rack servers—that offer supercomputing performance at a fraction of the cost and power consumption of CPU-only systems—is one way many are responding to this demand for increased productivity and faster computation abilities while maximizing ROI. Through Supermicro GPU rack server solutions, enterprises and institutions are able to leverage the capabilities of GPU computing via cost-efficient, easy-to-deploy form factors.

Supermicro GPU Rack Servers: High-Density, Enterprise-class Supercomputing Solutions

Optimized for compute-intensive applications ranging from oil and gas exploration to quantum chemistry, Supermicro’s high-performance GPU rack mount servers can provide your data center with the processing capacity and reliability you need to intelligently scale software performance and carry out mission-critical computing.

With support for both Kepler-based NVIDIA® Tesla® GPU computing accelerators and Intel® Xeon Phi™ coprocessors, Supermicro’s GPU-integrated SuperServers are both powerful and versatile, giving users a number of advantages including:

  • Performance. Unparalleled GPU density and Supermicro’s innovative, non-blocking GPU server architecture using PCI-E 3.0 expansion slots ensure optimal performance.
  • Efficiency and reliability. Redundant, platinum-level power supplies and intelligent cooling subsystems enhance the power efficiency and reliability of Supermicro GPU rack solutions.
  • Manageability. A built-in, IPMI 2.0-compliant server management utility allows you to monitor the health of and optimize cooling for your GPU rack server over a dedicated LAN connection.

Supermicro FatTwin™ GPU Rack Servers: Our Most Energy-efficient, Cost-Effective GPU Server Solutions

For enterprises looking to further “green” their data center or reduce their total cost of ownership while accelerating HPC applications, Thinkmate recommends our custom-configured Supermicro FatTwin GPU rack servers. Supporting up to 6 coprocessors per node, these 4U-twin systems are designed to achieve maximum GPU density, hard drive capacity, and power efficiency.

Compatible with industry-standard racks for easy, flexible deployment, FatTwin GPU servers extend the compute and storage capacities of Supermicro’s original twin SuperServer systems to boost server performance and storage space. And through the use of energy-saving system components and advanced cooling technologies, Supermicro was able to reduce server power consumption—making the FatTwin GPU rack exceptionally efficient. Through best-in-class resource optimization and a highly effective shared cooling architecture, which allows the system to operate at higher ambient temperatures, the FatTwin GPU rack offers both cost-effectiveness and reliability.

Leveraging cutting-edge GPUs to generate massively parallel processing power, Thinkmate-configured and supported Supermicro GPU servers are available in a variety of workload-optimized system configurations and other form factors including blades and Supermicro workstation solutions. Through our convenient online ordering system, you can find and optimize the right Supermicro GPU-enabled server or workstation for your specific applications.

Configure and Buy

Speak with an Expert Configurator at 1-800-371-1212