The Dell EMC DSS 8440 server is a 4U dual-socket system optimized with accelerators designed to deliver high compute performance for both training and inference. The system supports up to 10x double-wide or 16x single-wide NVIDIA GPUs, depending on the model. From healthcare to petroleum production, the system uses its computing horsepower to radically reduce the time for results and achieve new levels of efficiency and productivity.
As part of Dell’s GPU servers, the Dell EMC DSS 8440 server holds up to 24x DIMM slots, 10x drive bays, and up to 10x dual-width full-height PCI-Express slot graphics cards. The system also supports dual Intel Xeon Scalable processors from either the first or second generation and up to 768GB of memory.
The DSS 8440 is powered by two processors from either the first or second generation of Intel Xeon Scalable processors. Support for up to 24 cores per processor and a TDP up to 205W to help increase productivity for compute-intensive workloads like modeling, simulation, and predictive analytics. When configured with up to 10x NVIDIA V100 GPUs, 10x NVIDIA Quadro RTX GPUs, or 16x NVIDIA Tesla T4, the Dell 8440 delivers top-end compute performance needed to train even the most complex machine learning models.
Dell’s 8440 system features 12x module slots per CPU, for a total of 24x memory module slots with both processors. The server supports Registered (RDIMM) memory modules with a total memory of 768GB. All DIMM slots must be populated with the same memory module type.
At the rear of the chassis, the Dell 8440 has an array of 10x 2.5-inch hot-swap storage bays, which can be configured for NVMe, SAS, or SATA. The server supports different configurations such as 8x drive bays with support for NVMe and 2x bays with support for SAS/SATA, or vice versa.
The 4U server is optimized for expansion and offers PCI-Express based GPUs for deep learning and inferencing. The graphics cards are fitted on a printed circuit board (PCB) that handles PCIe slots separately from the motherboard. In addition, they sit towards the middle of the server, behind 12x 60mm x 60mm cooling fans, making it one of few GPU servers with this design. The server features a number of GPUs along with different memory sizes including:
Also, on the rear of the chassis users can find a selection of ports such as 2x 1GbE RJ45 ports plus 2x SFP+ ports plus a 4x daughter card.
The embedded Dell Remote Access Controller (iDRAC9) with Lifecycle Controller provides local and remote management of the system. Because of its multi-accelerator capacity, the Dell DSS 8440 makes it an ideal multi-tenancy solution. It can provide training and inference resources across multiple workloads, multiple users and departments, or multiple systems. Users have the flexibility to run different stacks of machine learning software on the same server using different numbers of accelerators. Multi-tenancy also lets data centers simplify the management of machine learning services using Kubernetes. Administrators can also leverage the NVIDIA GPU cloud and TensorRT Runtime Environment to allow data scientists, developers, and researchers to quickly create solutions, enhance performance, and add value to the company.
The Dell EMC DSS 8440 is a dual processor 4U server designed for high-performance computing, machine learning, and deep learning. With 24x DIMMs, 10x drive bays, and an extensive capacity for computational accelerator cards, this GPU server delivers results.
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.