The world’s first purpose-built system for deep learning is the NVIDIA DGX-1 which has fully integrated hardware and software meaning it can be deployed quick and easy. Its ground-breaking performance significantly accelerates training times, making the DGX-1 the first ever deep learning AI supercomputer in a box.
For deep learning success, data scientists and artificial intelligence researchers require accuracy, simplicity and speed, all these aspects are present in the DGX-1 allowing faster training and iteration which ultimately means faster innovation and a faster time to market.
NVIDIA Pascal is the groundbreaking architecture that is used in the NVIDIA DGX-1 - powered by the Tesla P100 accelerators with 8 of them interconnected using NVIDIA NVLink. Pascal was designed as the engine of computers that learn, see and simulate the real world, a world that has an infinate appetite for computing.
The NVIDIA DGX-1 software stack includes major deep learning frameworks, the NVIDIA DIGITS™ GPU training system, the NVIDIA Deep Learning SDK (e.g. CuDNN, NCCL), NVIDIA Docker, GPU drivers, and NVIDIA CUDA® for rapidly designing deep neural networks (DNN). It’s the ideal stack for accelerating popular analytics and visualization software.
This powerful DGX-1 includes access to cloud management services for container creation and deployment, system updates, and an application repository. The combination of these software capabilities running on Pascal-powered Tesla GPUs allows applications to run 12x faster than with previous GPU-accelerated solutions.
High-performance training accelerates your productivity, and when using the DGX-1 Supercomputer that means faster time to insight and faster time to market.
Note: Caffe benchmark with AlexNet, training 1.28M images with 90 epochs | CPU Server uses 2x Xeon E5-2697 v3 CPUs.
CPU is dual socket Intel Xeon E5-2698 v3. 170TF is half precision or FP16.
The turn key NVIDIA DGX-1 system provides plug-and-play setup and takes you from power-on to deep learning in minutes.
NVIDIA DGX-1 hardware and software support provides access to NVIDIA deep learning expertise and includes cloud management services, software upgrades and updates, and priority resolution of your critical issues.
Groundbreaking technologies allow the NVIDIA DGX-1 to deliver the fastest solutions for your deep learning training and AI-accelerated analytics workloads.
|Performance||170 teraFLOPs of half-precision (FP16) peak performance|
|GPU Architecture||8x NVIDIA Tesla P100 GPU accelerators based on the next-gen Pascal architecture|
|GPU RAM||16GB next-gen HBM2 memory per GPU|
|CPUs||2x Intel Xeon E5 2698 v4, 20 cores 2.2GHz|
|System RAM||512GB 2133MHz DDR4 ECC Registered|
|Storage||4x 1.92TB SSDs in RAID 0|
|Operating System||Ubuntu Server|
|Chassis||3U 19in rackmount chassis with 3200W PSU|
|CPU Type||Intel Xeon|
|No. of Threads||40|
|CPU Model||Xeon E5-2698 v4|
|Clock Speed||2.2 GHz|
|Burst Frequency||3.6 GHz|
|Memory Type (ECC)|
|Memory Type (R/U)|
|Memory Speed (Mhz)||DDR4 - 2133|
|Max. Memory Capacity|
|On Board VGA||No|
|Storage Drive Included||Yes|
|Storage Drive Type|
|Storage Drive Size||4x 1920 GB|
|Processor Graphics||8x NVIDIA Tesla P100|
|Base GPU Speed||1328 MHz|
|Max. GPU Speed||1480 MHz|
|Max. GPU Resolution|
|Keyboard / Mouse||No|
|Operating System Included||Yes|
|Operating System||Ubuntu Server Linux OS|
|Dimensions||866 x 444 x 131 (mm WxHxD)|
Please note your statutory rights are not affected.
For further information regarding Scan's warranty procedure please see our terms and conditions