Use Nvidia’s DGX-1 Supercomputer To Build Your Own Skynet

nvidia-dgx1-deep-learning-supercomputer-1

It may look like any box you see lining up the shelves in the server room at work, but there’s nothing ordinary about the Nvidia DGX-1.  Housed inside that enclosure is a GPU-accelerated supercomputer that delivers the throughput equivalent of 250 x86-based CPU servers.

Designed to handle the processing requirements of deep learning and artificial intelligence, the system boasts up to 170 teraflops of half-precision at peak performance, allowing it to train deep neural networks at a rate 12 times faster than Nvidia’s Maxwell architecture from just last year. That means, whether you’re on working on self-driving cars, autonomous drones, or your very own version of Skynet, this thing should help your AI systems realize their potential like no other off-the-shelf server box can.

nvidia-dgx1-deep-learning-supercomputer-2

The Nvidia DGX-1 is a turnkey system that combines eight 16GB Tesla P100 GPUs, a pair of Intel Xeon CPUs, 512GB of RAM, four 1.92TB SSDs, dual 10GbE Quad InfiniBand 100Gb networking, and an NVLink Hybrid Cube Mesh, all cobbled up together in a single box. Each one comes preloaded with a suite of software optimized for deep learning, including the NVIDIA DIGITS (a deep learning training system), the NVIDIA CUDA (a library of primitives for building deep neural networks), and optimized versions of deep learning frameworks, such as Caffe, Theano and Torch.  As you can imagine, this is no lightweight box, tipping the scales at a hefty 132 pounds.

No release date has been given, but the Nvidia DGX-1 is priced at $129,000.

Check It Out

2 Responses

  1. Jo

    This is absolutely amazing. Now, I’m not going to sit here and pretend like I would know how to use technology like this myself or that 129 grand is anywhere near any sort of budget I currently have but what I wouldn’t give to be part of creating products such as this. I find it so fascinating that artificial intelligence has come so far and that such integrate software are readily available. Sounds like this system is still going through some finishing touches, but I can’t wait to learn more and see this bad boy in action. Man… what’s next??

    Reply
  2. Pete

    So this is an advanced server designed to help increase the efficiency of artificial intelligence? I don’t understand how this works. Is it just to relay data to the AI device? I do understand servers though. My website is on a VPS and has 2 GB RAM and a 60GB SSD Drive. This thing has 512 GB of RAM and 4 1.92 Terabytes of SSD storage. Holy crap! I can assure you this thing will cost more than my home. That is crazy. The cache alone is 7 TBs!

    I definitely need to read more on the NVIDIA DIGITS because this looks like a fascinating topic to nerd out on. If you look at what this computer has spec-wise, it would have been impossible to have something like this in the 80’s or even the 90’s. This is some truly futuristic machinery here, and ANYONE can buy it, as long as you have probably $150,000 to spend.

    If I could afford to waste money, I’d love to buy one and play around with it. It might make a good investment to buy one and then rent it out to researchers. Maybe not, but either way this is one magnificent server. The stats on this is astounding. I wonder how many of these they expect to sell each year? What will their price mark-up be? How long until this servers stats are outdated and old. I am excited to see what else the tech world will come out with over the next 5-10 years.

    Reply

Leave a Reply

Your email address will not be published.