NVIDIA DGX Station
Use Your Own Personal AI Supercomputer to Dive into Deep Learning – Cutting-Edge AI at Your Desktop
The NVIDIA DGX Station is the world’s very first personal supercomputer – built specifically for bleeding-edge AI development. It uses the same Cloud Deep Learning Stack that powers all NVIDIA DGX systems – allowing you to experiment both at your desk and across all DBX-based cloud systems. This allows for easy, simplified experimentation from your desk to your NVIDIA DGX Station – and everywhere in between.
DGX Station vs. DIY
For specification details
The Only Supercomputer Designed for Your Office
Whisper-quiet, with a beautifully crafted design and incredible performance, the DGX Station by NVIDIA is designed to help you experiment with AI in your office, lab, or workspace.
Get the Fastest Start in Deep Learning
Your work should be in discovering new AI applications – not hardware setup. With the DGX Station, you can minimize setup and deployment time, and avoid deployments that take weeks. Deployment is both intuitive and simple. All you need to do is plug it in and power it up. With an integrated hardware and software design, NVIDIA DGX Station empowers you to spend more time collecting insights and doing meaningful work – and less time on setup.
Productivity from your Desk to the Data Center
Stop spending time and money on IT when you could be spending it on data science. The DGX Station can save you quite a bit of money – both in downtime and inefficiency when designing open-source frameworks.
With the DGX Station, you can start to take the experimentation to your desk – and scale it to data-center sized workloads using the NVIDIA DGX-1, or the cloud, using NVIDIA GPU Cloud Deep Learning technology stacks and containers. Our goal with the NVIDIA DGX Station is to enable you to complete tasks unhindered by technological constraints.
4X Faster Than the Fastest Workstations
With the computing capacity of a four-rack server in a single, desk-friendly package, the DGX Station provides incredible performance – at less than 1/20th the power consumption of a comparable server. Using NVIDIA NVLink technology, Tesla V100 GPUs, and a brand-new Tensor Core architecture, the DGX Station provides 500 TFLOPS of power, and a water-cooled design ensures that it stays cool and efficient while providing 3x the performance of the most cutting-edge workstations on the market.