Part of Supermicro’s GPU server line, the Supermicro SuperServer 1029GQ-TRT-NEBS supports dual Intel Xeon Scalable processors, 4x NVIDIA GPUs, and 12x memory module slots. The features in this 1U rackmount system are designed for 5G, cloud computing, telecom central office and data center, AI inference and machine learning, as well as network functions virtualization.
The Supermicro SuperServer 1029GQ-TRT-NEBS can support up to 4x NVIDIA Tesla V100 GPUs and up to 4TB of memory when using Intel Optane DC persistent (DCPMM) memory modules. The storage on this system is flexible with four drive bays and an M.2 slot to support an SSD device. The system is optimized for high-performance computing and is ideal for data center applications.
At the heart of the 1029GQ-TRT-NEBS, there are dual Intel Xeon Scalable processors from either the first or second generation. When using 2nd generation CPUs, memory speeds can reach up to 2933MHz and Intel Optane DC persistent (DCPMM) memory modules are supported. This platform will accommodate Intel Xeon Scalable processors up to 28 cores per socket with a maximum TDP of 165W. The UltraPath Interconnect (UPI) is the communication channel between CPUs and supports speeds of up to 10.4 GT/s. The 1U server is powered by redundant 2000W level power supplies certified 80Plus Titanium level, which is the highest efficiency rating.
With 6x memory channels per CPU, the Supermicro X11 generation motherboard has 6x module slots per CPU, for a total of 12x. The server supports up to 3TB of Load-Reduced of 3DS ECC DDR4 (LRDIMM) or Registered (RDIMM) memory modules. Using Intel Optane DC persistent (DCPMM) memory modules, up to 4TB of memory can be installed. Memory speeds can reach 2933MHz if both CPU and memory modules support that speed, defaulting to the slower performer between the two.
The server supports up to four drive bays. The front of the SuperServer 1029GQ-TRT-NEBS has 2x 2.5-inch hot-swap bays that support SAS or SATA. Similarly, there are two more internal 2.5-inch drive bays with support for SATA. Also under the hood, the motherboard has one M.2 slot designed to support a single SSD device in a PCI-Express 3.0 NVMe or SATA interface for native support.
The Supermicro 1029GQ-TRT-NEBS supports up to four NVIDIA Tesla V100 GPUs nestled behind seven fans pulling fresh air from the front of the chassis. To accommodate the GPUs, the motherboard features 4x PCIe 3.0 full-height, full-length slots plus 2x PCIe 3.0 low-profile slots. On the rear of the chassis, there’s a selection of ports including 2x RJ45 10GbE ports and a dedicated IPMI LAN port. There’s also a VGA port, dual USB 3.0 ports, dual PCI-e 3.0 x16 slots, and a space for a GPU. The 1U server is powered by redundant 2000W level power supplies. This server is certified 80Plus at the Titanium level to promote highly-efficient power supplies.
There are several management features to control the server remotely and easily. The SuperServer 1029GQ-TRT-NEBS supports Supermicro SuperDoctor 5 in a command-line or web-based interface. The program monitors system health like processor temperature, system voltages, and fan speed. The system also offers support for Watch Dog and the Intelligent Platform Management (IPMI 2.0) which can similarly be used to provide remote access, monitoring, and management.
The Supermicro SuperServer 1029GQ-TRT-NEBS is a dense server packed into a 1U form factor. It’s equipped with dual Intel Xeon Scalable processors from either the first or second generation, up to 12x DIMMs, and 4x NVIDIA GPUs. The storage on this system is flexible with two drive bays on the front, two drives on the back, and a M.2 slot for SSD devices. The SuperServer can easily be scaled for future growth of virtualization organizations.
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.