Nvidia has unveiled a supercomputer designed specifically for deep learning to meet the demands of artificial intelligence (AI).
The Nvidia DGX-1 supercomputer is claimed to deliver a throughput of 250 x 86 servers by combining a new generation of Nvidia GPU accelerators, deep learning software and development tools.
It allows data scientists and researchers to create a new class of intelligent machines that learn, see and perceive the world as humans do, Nvidia said.
The computer’s performance significantly accelerates training time, making it the world’s first deep learning supercomputer in a box.
The DGX-1 is built with Nvidia Pascal- powered Tesla P100 accelerators, interconnected with Nvidia NVLink for high-speed interconnect for maximum application scalability; 16nm FinFET fabrication technology for energy efficiency.
It also includes Chip on Wafer on Substrate with HBM2 for big data workloads and new half-precision instructions will provide over 21 teraflops of peak performance for deep learning.
The DGX-1 software houses the Nvidia Deep Learning GPU Training System (DIGITS), an interactive system used to design deep neural networks (DNNs).
It also features the Nvidia CUDA Deep Neural Network library (cuDNN) version 5, a GPU-accelerated library of primitives for designing DNNs.
The supercomputer offers access to cloud management services for container creation and deployment, system updates, and an application repository.
The combination of all the software capabilities running on Pascal-powered Tesla GPUs enables applications to run 12 times faster than with earlier GPU-accelerated solutions.
Nvidia CEO and co-founder Jen-Hsun Huang said "Artificial intelligence is the most far-reaching technological advancement in our lifetime.
"Data scientists and AI researchers today spend far too much time on home-brewed high performance computing solutions. The DGX-1 is easy to deploy and was created for one purpose: to unlock the powers of superhuman capabilities and apply them to problems that were once unsolvable."
The DGX-1 will be available in the US in June, and in other countries during the third quarter of this year.
Nvidia said its Pascal architecture has strong support from the artificial intelligence ecosystem.
Microsoft Research chief speech scientist Xuedong Huang said: "Microsoft is developing super deep neural networks that are more than 1,000 layers.
"Nvidia Tesla P100’s impressive horsepower will enable Microsoft’s CNTK to accelerate AI breakthroughs."
