I’ve already posted about it here and here on the blog: Artificial intelligence will be one of the dominant topics in the coming years. According to research from Gartner, within the decade AI will develop into a mass trend. Another study by IDG Research Services, found that companies classify machine learning (ML) and AI as key technologies.
This opens up a huge market for companies. But what many of them underestimate is that successful AI projects are based on the ability to deliver and manage huge amounts of data.
We wanted to find a solution. As NetApp’s focus is on data management in the hybrid cloud, we know how important it is that data can be stored securely wherever it’s needed. Data needs to be managed easily and must be able to flow quickly and securely. Deep learning in particular uses and analyzes huge amounts of data. This requires special computers. This is where NVIDIA came into play with its DGX supercomputer.
So we approached NVIDIA with our idea: How about combining one of the world’s strongest GPU solutions with our fast cloud-integratable All Flash Storage AFF A800 to create an AI architecture? Almost half a year ago, we introduced this solution in partnership with NVIDIA. The request resulted in NetApp ONTAP AI Proven Architecture. The key feature of this solution is that it simplifies and accelerates the movement of data between the point of origin, the data center, and the cloud – thereby accelerating the implementation of AI projects. We offer an architecture where customers can run their large and small AI projects. This is because the architecture is easily scalable, so that companies no longer have to worry about how they manage their data, where they need it, or how they can protect or duplicate it. That is what we do with our Data Fabric approach.
However, our collaboration with NVIDIA doesn’t end with joint development. For example, we provide a joint technical support network for ONTAP AI Proven Architecture.
Who can benefit from our architecture
Who is the solution suitable for? As I said before, with ONTAP AI Proven Architecture and our Data Fabric, we provide companies with a proven architecture that is scalable. This allows users to leverage data for AI projects or scale applications as required by the project. So they can concentrate on the essentials instead of having to deal with the underlying infrastructure.
The example of companies dealing with autonomous driving is particularly vivid. In the optimal case, AI controls the cars by using self-learning algorithms. It constantly evaluates a huge amount of data, which is collected by numerous sensors, for example. The large amount of data has to be transported from the cars to the data centers, where it is processed in various stages with the help of deep learning applications. So far, different infrastructures have been used for each step, such as data recording, testing or archiving. This costs a lot of time and slows down productivity. The ONTAP AI Proven Architecture provides a single infrastructure for all tasks in a single data pipeline. This allows much more data to be evaluated at the same time. Equally enormous amounts of data are processed in the development of AI concepts in the fields of health or consumer electronics. The projects of the British company Cambridge Consultants give us an idea of what we can look forward to in the future. Using NetApp technology, for example, they run the artist Vincent: this AI combines the skills of Picasso and Van Gogh to complete paintings that people began as sketches.
What will be possible in the future cannot be foreseen today. We are looking forward to the next six months with ONTAP AI Proven Architecture!