Back in 2018, NVIDIA DGX Station was a new comer in the world of supercomputer. Built with towering physical, it still manages to be desk-friendly and amazingly quite in nature due to its water-cooled internal features. It is put in the same class as its predecessor, DGX-1 right away, but in slightly more accessible ways both in deployment and cost factor. This supercomputer is the little siblings of DGX-1 in various ways possible. It has half the power draw, half the system memory, half the CPU, half in GPU. The NVIDIA DGX Station price is also almost half of the DGX-1.
What Factors Justify NVIDIA DGX Station Price?
Even though NVIDIA DGX Station is designed to be the down-sized version of NVIDIA DGX-1 server, it is still able to hold the performances on its own just as fine. It comes with loaded amounts of performance-boosting additives. The DGX Station has three 1.92TB SSDs for the operating system and in RAID 0, it has 256GB of DDR4 memory, and dual 10GB LAN solution.
Now while the manufacturer is able to cut down NVIDIA DGX Station price to half of its brother’s, the enthralling part is the DGX Station in fact carries more compute power than the superior models. NVIDIA DGX-1 has 170 TFLOPs and the DGX Station has 480 TFLOPS. It is all possible due to the new compute unit that based on Tesla V100. If this fact has successfully piqued your interest already, check out several qualities of DGX Station listed below.
1. Fastest Deep Learning Start
Rather than discovering new hardware setup, it is the era of discovering new AI application. You will be able to cut the deployment time that sometimes take several weeks and minimize setup with the DGX Station. All you need to do is just plug the device in and power it up, the rest goes in simple and intuitive ways. It also incorporated integrated design of hardware and software so you can spend less time on setup and save it for collecting some insights.
2. Increasing productivity to submit to Data Center
Let’s be honest, the time and money spent on IT can be spend better on data science. By using DGX Station, there is chance to safe both money and down time to encourage efficiency while designing frameworks from open sources. You’d be able to experiment at your desk, scaling data to the Data Center that uses NVIDIA DGX-1 or to the NVIDIA GPU Cloud Deep Learning technology.
3. 4 times faster than the fastest
DGX Station offers highly efficient performance. It is designed to be desk-friendly, yet has capacity of computing by combining four rack servers in one. The power consumption goes as low as 1/20th of comparable server’s consumption of power. Considering the NVIDIA DGX Station price, the 500 TFLOPS and water-cooled design is very appreciated. Those features are possible due to the use of new Tensor Core architecture, Tesla V100 GPUs and NVIDIA NVLink technology. The machine is able to stays cool and work efficiently while producing three times the performance of available workstations in the market.
NVIDIA DGX Station is an economical yet powerful choice of supercomputer. It built specifically to be the forefront of AI development. It is designed to be personal as well, as the size allows you to experiment the experience at your desk and across the cloud systems with DBX bases. The NVIDIA DGX Station price might seem a little bit extreme at first, but considering that this beautifully-crafted supercomputer provide exceptional performance and offers simplified experimentation with AI in your workspace, lab, or office, it might be worth to get.