NVIDIA leans on AI, deep learning to stay relevant for 4th industrial wave

The California based company is finding new ways to stay relevant in the AI driven technology space. Working at the cutting edge of deep learning, the company is optimistic about Chennai and its start-ups working in areas like IoT, AI, machine learning and more. 

In an interview, Vishal Dhupar, MD, South Asia, NVIDIA Graphics Pvt Ltd, talks about the tech giant’s focus on deep learning and the promising fields going forth.

The next Olympics in Japan could feature autonomous taxis that can be hired on your mobile. And Netherlands is also talking about auto taxis seriously. AI is being hailed as the fourth wave of the industrial revolution and is slated to work at 20 times the scale of any industrial revolution that has taken place so far. It’s a trillion-dollar industry on many counts including auto, healthcare, retail, speech and language translation.

Deep Learning takes a look at data and infers the intricate feature sets inside that and understands how to obtain knowledge from them, and through that predict, observe and infuse intelligence into devices. NVIDIA has been a great proponent of deep learning. The work that we have done for over a decade and a half, makes us a platform of choice for it. Instead of resting on our laurels, we want to ensure that in the early stages before this computing model is understood by everyone and adopted, people are successful while using it. 

This is a complete transformation for NVIDIA, as predominantly before that we were a component to somebody else’s system. Now we are a platform for this part. India was also a shining star when it came to writing software, but what if software has to write software? That requires a different technique. And if we don’t quickly transform, somebody else might take advantage of that. If deep learning can be learned, applied and used by end businesses, that will be a good achievement for us. Our company has single-mindedly been focussed on accelerated computing. The first application we observed for high performance computing application was in computer graphics. Gaming had a heavy duty computational requirement as predominantly, gamers needed to be brought closer to simulated reality. You had to change the paradigm of how the technology can come down to the price point of end users or consumers. The second area we concentrated on was high performance computing – for weather simulation, material science, fluid dynamics. Domain by domain we tried to solve the challenges of the computational process. That’s where this processor, which we named as GPU and the CPU would come together and increase the velocity of the science, art and craftsmanship.

As it was two different DNAs – one very good for sequential programming, the other for parallel programming, the arrival of deep learning is inherently parallel. So, we decided to work right from the foundation of computing, with the operating system, with the middleware, application teams and ensure they had a platform for development. 

Chennai centric 

We created a six city developer connect programme in India to train people on deep learning and practical application of the same through our institute and tell them how to work on several use cases envisaged by us. Some of the key platforms we have built are for autonomous vehicles, IoT. I was excited with Chennai. It is a place of promise and intellect. AI can be a differentiator as there is a huge industrial framework here. If AI is infused into industry, it would create a sea change for the industry and its workers.