Nvidia to Launch Enterprise AI Data Platform Ontap AI

The basis of the platform is the supercomputers DGX-1 and cloud flash storages AFF A800
03 August 2018   799

NVIDIA and NetApp launched the corporate platform Ontap AI to store models of artificial intelligence. Unlike cloud services, it is equipped with tools for accelerated collection, processing and transmission of data. This is reported by Venture App.

The basis of the platform is the supercomputers DGX-1 and cloud flash storages AFF A800. To distribute data over the clouds and provide quick access, regardless of format and location, the Data Fabric architecture is used.

The NVIDIA DGX-1 computers support second-generation in-depth training and are equipped with Tesla V100 graphics cards. One DGX-1 rack provides 1 Pflops of power and is able to train the FairSeq Neural Network presented in May 2017 for one and a half days. The AFF A800 drives in a cluster with 24 nodes read data at 300 GB / s and have a delay of 200 μs.

The Ontap AI platform was already used in the consulting firm Cambridge Consultants. It was involved in the development of systems for studying the effects of drugs on patients. Also Ontap AI was used to create Vincent - a painting training program at the level of human capabilities.

In April 2018, NVIDIA CEO Jensen Huang announced that their video cards no longer comply with Moore's law. With the help of the DGX-2 supercomputer, the company managed to train the AlexNet neural network to store 15 million images in 18 minutes.

Neural Network to Create Landscapes from Sketches

Nvidia created GauGAN model that uses generative-competitive neural networks to process segmented images and create beautiful landscapes from peoples' sketches
20 March 2019   127

At the GTC 2019 conference, NVIDIA presented a demo version of the GauGAN neural network, which can turn sketchy drawings into photorealistic images.

The GauGAN model, named after the famous artist Paul Gauguin, uses generative-competitive neural networks to process segmented images. The generator creates an image and transfers it to the discriminator trained in real photographs. He in turn pixel-by-pixel tells the generator what to fix and where.

Simply put, the principle of the neural network is similar to the coloring of the coloring, but instead of children's drawings, it produces beautiful landscapes. Its creators emphasize that it does not just glue pieces of images, but generates unique ones, like a real artist.

Among other things, the neural network is able to imitate the styles of various artists and change the times of the day and year in the image. It also generates realistic reflections on water surfaces, such as ponds and rivers.

So far, GauGAN is configured to work with landscapes, but the neural network architecture allows us to train it to create urban images as well. The source text of the report in PDF is available here.

GauGAN can be useful to both architects and city planners, and landscape designers with game developers. An AI that understands what the real world looks like will simplify the implementation of their ideas and help you quickly change them. Soon the neural network will be available on the AI ​​Playground.