The Netstor 4U Rack Mount Expansion Chassis with 1000W Single Power Supply from Dynapower USA is a professional 19" rack mount PCIe Gen3 GPU enclosure that features ultra-high flexibility to expand the I/O capabilities of a host PC, server or workstation. It provides four PCIe 3.0 ×16 slots and two PCIe 3.0 ×4 slots (optional) on backplane for accommodating all combinations of PCIe Gen3 ×1, ×4, ×8 and ×16 cards or multiple high-performance GPU cards, while maintaining backwards compatibility with PCIe Gen2 applications in the HPC field. The enclosure turns a server or workstation into a supercomputer for running GPU-accelerated programs, dramatically enhancing performance up to ten times beyond the capabilities of the present-day multi-core CPU technology.
The professional enclosure features PCIe 3.0 × 16 host cards and a PCIe 3.0 data cable, providing up to 128 Gbps extremely high-speed data transfer rates between the host and GPU enclosure for high-performance applications in supercomputing, simulation and video post-production. It allows multiple PCIe Gen3 GPU cards, RAID controller cards, FPGA cards and audio/video accelerator cards to work with the server or workstation. The enclosure is designed to work with mission-critical tasks in the fields of high-performance computing, telecommunication, data analysis, surveillance, storage and so on. This enclosure provides dedicated sufficient power supply with additional eight 6+2 pin PCIe power connectors particularly for up to 4 double-width GPU cards for HPC. The patented, dedicated ventilation design and big cooling fans offer excellent ventilation for GPUs, totally relieving the heat loading of server or workstation. With this PCIe GPU expansion rack mountable enclosure, even the space-limited 1U or 2U server can be transformed into a high-performance computer for applications including digital content creation, medical research, financial simulation, 3D seismic interpretation, gene sequencing and more. Work is performed faster and more efficiently, meaning quick results as well as increased energy savings compared to running on CPUs alone.