NVIDIA DGX Systeme

The world's most powerful AI system for the most complex AI challenges

Break through the barriers to AI speed and scale with NVIDIA DGX-2, the first 2 petaFLOPS system that engages 16 fully interconnected GPUs for 10X the deep learning performance. Powered by NVIDIA® DGX™ software and an architecture designed for AI-scale built on NVIDIA NVSwitch, DGX-2 enables you to take on the world’s most complex deep learning challenges.

Powered by NGC Deep Learning Stack

• Integrated suite of optimized deep learning software
• Simplified workload management
• Free download: NVIDIA DGX systems deep learning software brief

• Get started in one day instead of months
• Simply unpack, plug-in, and start getting results

• Save hundreds of thousands of dollars in engineering effort
• Avoid months of lost productivity spent on IT

• NVIDIA DGX Systems with Volta offers 3X the speed of prior generations
• Deep learning training, inference and accelerated analytics in one system

The World's First Deep Learning Supercomputer in a box

The NVIDIA® DGX-1™ is the world’s first purpose-built system for deep learning with fully integrated hardware and software that can be deployed quickly and easily. Its revolutionary performance significantly accelerates training time, making the NVIDIA DGX-1 the world’s first deep learning supercomputer in a box.

The Personal Supercomputer for Leading-Edge AI Development

Your data science team depends on computing performance to gain insights, and innovate faster through the power of deep learning and data analytics. Until now, AI supercomputing was confined to the data center, limiting the experimentation needed to develop and test deep neural networks prior to training at scale. Now there’s a solution, offering the power to experiment with deep learning while bringing AI supercomputing performance within arm’s reach.

Kurzbeschreibung:

• Up to 8 Tesla P100 SXM2
• Up to 80 GB/s GPU-to-GPU NVLINK
• Supports GPUDirect RDMA

1. Dual socket R3 (LGA 2011) supports Intel® Xeon® processor E5-2600 v4† /v3 family; QPI up to 9.6GT/s
2. Up to3TB†ECC 3DS LRDIMM, up to DDR4- 2400† MHz; 24x DIMM slots
3. 4 PCI-E 3.0 x16 (double-width) slots 2 PCI-E 3.0 x8 slots
4. Dual 10GBase-T LAN with Intel® X540
5. 16 Hot-swap 2.5" SATA/SAS drives; Supports 8 NVMe drives
6. 8x 92mm cooling fans
7. 2000W Redundant (2+2) Power Supplies;Titanium Level (96%+)

sysGen/SUPERMICRO Server SYS-4028GR-TVRT, bestehend aus:
Mainboard MBD-X10DGO-T / 4U Rackmount CSE-R422BG

• Up to 8 Tesla V100 SXM2
• Up to 300 GB/s GPU-to-GPU NVLINK
• Supports GPUDirect RDMA

1. Dual socket R3 (LGA 2011) supports Intel® Xeon® processor E5-2600 v4† /v3 family; QPI up to 9.6GT/s
2. Up to 3TB†ECC 3DS LRDIMM, up to 3TB DDR4- 2400† MHz; 24x DIMM slots
3. 4 PCI-E 3.0 x16 (double-width) slots 2 PCI-E 3.0 x8 slots
4. Dual 10GBase-T LAN with Intel® X540
5. 16 Hot-swap 2.5" SATA/SAS drives; Supports 8 NVMe drives
6. 8x 92mm cooling fans
7. 2000W Redundant (2+2) Power Supplies;Titanium Level (96%+)

• Artificial Intelligence
• Big Data Analytics
• High-performance Computing
• Research Lab/National Lab
• Astrophysics, Business Intelligence

1. Dual socket P (LGA 3647) supports Intel® Xeon® Scalable Processors, 3 UPI up to 10.4GT/s
2. Up to 3TB ECC 3DS LRDIMM, up to DDR4-2666MHz; 24 DIMM slots
3. 4 PCI-E 3.0 x16 (LP) (GPU tray for GPUDirect RDMA), 2 PCI-E 3.0 x16 (LP, CPU tray)
4. Support 8 Tesla V100 with 300GB/s NVLINK
5. Dual 10GBase-T LAN with Intel® X540
6. 16 Hot-swap 2.5" SAS/SATA drives; Supports 8 NVMe drives
7. 8x 92mm cooling fans
8. 2200W Redundant (2+2) Power Supplies; Titanium Level (96%+)