The Tyan Thunder HX FA77B7119 B7119F77V4HR-2T-N Server is a high-performance platform that delivers an impressive array of technology for addressing Artificial intelligence, Deep Neural Networks, and a myriad of other massively parallel workloads. It supports up to 10 GPGPUs in a 4U form factor, supported by dual Intel Xeon Scalable processors, and up to 3TB of memory.
With support for up to 10 GPUs, the Tyan Thunder HX FA77B7119 B7119F77V4HR-2T-N Server is designed for GPUDirect Remote Direct Memory Access (RDMA) applications, Deep Neural Networks, and AI. RDMA improves performance with ridiculously parallel workloads enabling faster data transfers and processing, especially when both platforms support RDMA. Deep Neural Networks and Artificial Intelligence emulate the way the human brain functions, and attempt to model computer interactions after human analogues. GPU-enhanced servers are the preferred processing platform for data science plus AI research and implementation. What defines this barebones chassis is Tyan’s S7119GMRE motherboard. Other configurations based on this motherboard support more storage, including NVMe drives.
Tyan Thunder HX B7119F77V14HR-2T-N Server 14 SFF SATA 6Gb/s
Tyan Thunder HX FA77B7119 B7119F77V10E4HR-2T-N Server 10 SFF SATA 6Gb/s + 4 SFF NVME U.2 Bays
Tyan Thunder HX FA77B7119 B7119F77V14HR-2T55-N
14x 2.5" SATA 6G Hotswap Bays
Tyan Thunder HX FA77B7119 B7119F77V10E4HR-2T55-N
10x 2.5" SATA 6G + (4) 2.5" NVMe U.2 Hotswap Bays
Tyan Thunder HX FA77B7119 B7119F77V4HR-2T55-N
4x 3.5"/2.5" SATA 6G Hotswap Bays
Dual 1st or 2nd Gen Intel Xeon Scalable processors with a Thermal Design Point (TDP) of up to 205W are supported. Augmenting performance, are up to ten Nvidia Tesla GPUs. The system is powered by up to four 4,800W hot-swap 80-plus platinum PSUs in a 3+1 redundancy configuration. Optional network communications include support for data transfer speeds of up to 100Gb/s.
Up to four 3.5-inch hot-swap SAS or SATA HDDs or SSDs are supported on this system offering 6Gb/s natively, and 12Gb/s access speeds with a separate HD controller installed. Both 3.5-inch and 2.5-inch drives can be installed on the Tyan FA77B7119 B7119F77V4HR-2T-N Server but 2.5-inch drives will require a drive tray converter. High-performance HD/RAID controllers offer a full range of RAID configurations supporting performance and resiliency.
Each processor supports six memory channels with two slots each for a total of 12 DDR4 DIMM slots per processor. There are 24 memory module slots total in a two-processor configuration. Using 3DS load-reduced (LRDIMM) or registered (RDIMM) memory modules will provide the greatest capacity at up to 3TB. Using registered DIMM modules, the system will support a maximum of 768GB, and with load-reduced memory modules you’ll get up to 1.5TB. Only the 3DS memory modules will support the full 3TB capacity when paired with a Xeon Scalable processor with an “M” after the product number. Other Scalable processors without the “M” only support a maximum of 768GB of memory, each.
There are 11 x PCIe 3.0 x16 slots on the motherboard, 10 of which support full-height double-width Tesla GPUs from Nvidia. The eleventh slot is dedicated for a high-performance I/O card. On the back of the system are two integrated 10GbE ports for network connectivity and a single GbE port dedicated to the Intelligent Platform Management Interface (IPMI). Optional storage mezzanine cards support 12Gb/s access speeds plus more RAID options.
The Onboard Base Management Controller (BMC) with IPMI 2.0 ensures ease of management for your system. The MegaRAC GUI grants remote control over your system, which provides a simple and effective way to monitor your system from any location. The control panel at the front of the system houses LED lights to indicate the status of the server and its general health.
The Tyan Thunder HX FA77B7119 B7119F77V4HR-2T-N Server has many uses ranging from Artificial Intelligence to GPUDirect Remote Direct Memory Access applications. This system is a must for running massively parallel workloads with additional processing power provided by up to 10 high-performance Tesla computational accelerators.
If you know what you want but can't find the exact configuration you're looking for, have one of our knowledgeable sales staff contact you. Give us a list of the components you would like to incorporate into the system, and the quantities, if more than one. We will get back to you immediately with an official quote.