Deep Learning & AI - Training

This powerful Deep Learning training system is a 2U rackmount server capable of supporting up to 4 PCI-E dual-slot NVIDIA GPU accelerators with a host CPU from the AMD EPYC processor series with up to 64 cores and 1TB of 8-channel DDR4 memory.

This next-generation server is based on the Intel Xeon Scalable processor family which has more PCI-E lanes and 50% memory bandwidth than the previous generation Xeon E5 2600 v4 processors. With support for up to four NVIDIA GPU accelerators and multiple storage bays this server provides great flexbility in a compact 2U form factor.

This powerful Deep Learning training system is a 2U rackmount server capable of supporting up to 3 PCI-E dual-slot NVIDIA GPU with a pair of host CPUs from the AMD EPYC processor series with up to 64 cores and 4TB of 8-channel DDR4 memory.

This powerful Deep Learning training system is a 2U rackmount server capable of supporting up to 8 PCI-E dual-slot NVIDIA GPU accelerators with a host CPU from the AMD EPYC processor series with up to 64 cores and 1TB of 8-channel DDR4 memory.

The Deep Learning GI8000-TR3 is a 4U rackmount server capable of supporting up to 8 PCI-E dual-slot NVIDIA GPU accelerators from the Tesla product family plus a pair of powerful Intel Xeon Scalable processors, each with up 28-cores, plus up to 1.5TB of 6-channel DDR4 memory.

This powerful Deep Learning training system is a 2U rackmount server capable of supporting up to 8 PCI-E dual-slot NVIDIA GPU accelerators with two host CPUs from the AMD EPYC processor series each with up to 64 cores and 2TB of 8-channel DDR4 memory.

The Deep Learning GI8000-AR1 is a 4U rackmount server capable of supporting up to 8 PCI-E dual-slot NVIDIA GPU accelerators plus a pair of powerful Intel Xeon Scalable processors, each with up 28-cores, plus up to 1.5TB of 6-channel DDR4 memory. It has a unique programmable PLX architecture that can either share the 8 GPUs between the two CPUs equally, or connect all 8 GPUs via a single CPU for improved performance in certain Deep Learning workloads.

This powerful Deep Learning training system is a 4U rackmount server capable of supporting up to 8 PCI-E dual-slot NVIDIA GPU accelerators from the Tesla product family, with two hosts CPUs from the AMD EPYC processor series with up to 64 cores and 4TB of 8-channel DDR4 memory.

The Deep Learning GI8000-GR2 is a 4U rackmount server which supports eight NVIDIA Tesla GPUs which communicate via NVLink which provides much higher bandwidth than PCI-E. The host CPUs are a pair of powerful Intel Xeon Scalable processors, each with up 28-cores, plus up to 1.5TB of 6-channel DDR4 memory.

Introducing the world's first Deep Learning Supercomputer, the DGX-1. Powered by eight NVIDIA Tesla V100 GPU accelerators specifically built for deep learning and machine learning, the DGX-1 provides extremely fast processing so you can explore multiple network architectures and manage and collect datasets to speed up the delivery of data. Education pricing available on request

The A100 is the latest generation NVIDIA DGX appliance and is powered by eight Tesla A100 GPUs that provide a combined 5 petaFLOPs of performance. The new architecture also benefits from improved NVLink and NVSwitch interconnects to boost performance in ddep learning and machine learning training. Education pricing is available on request.

The second generation DGX-2 provides double the compute power thanks to having 16 NVIDIA Tesla V100 GPU accelerators specifically built for deep learning and machine learning. The DGX-2 provides the fastest possible research so you can explore multiple network architectures and manage and collect datasets to speed up the delivery of data. Education pricing available on request

This next-generation server features two Intel Xeon Scalable processors, which have more cores and 50% extra memory bandwidth than the previous generation Xeon E5 2600 v4 processors for excellent performance in Deep Learning and Machine Learning training.

This next-generation server features four dual-socket nodes for a total of eight Intel Xeon Scalable processors, which have more cores and 50% extra memory bandwidth than the previous generation Xeon E5 2600 v4 processors for excellent performance in Deep Learning and Machine Learning training.