AI computing may have become a buzzword over the last decade or so, but its seeds were sown by innovators like NVIDIA much before. After being the pioneers in the introduction of Graphics Processing Unit (GPU), the company is at the forefront today in tackling challenges that no one else can solve. No wonder, their work in AI and the metaverse is revolutionizing several industries, one step at a time.

For over a decade now, Vishal Dhupar – Managing Director, South Asia has been responsible for running all aspects of NVIDIA’s business operations. His role involves developing and maintaining relations with the government and public institutions, as well as industry representative bodies.

With over three decades of experience in the information and communication technology industry, Dhupar has managed and led companies across diverse businesses – servers, graphics, digital equipment, data storage, and more. Previously, he has been credited with the success of companies like Symantec, Autodesk Inc, Sun Microsystems, and Silicon Graphics Inc, among others.

In a detailed chat with Jibu Elias — Content and Research Head of INDIAai, Dhupar delves deep into accelerated computing, the power of generative AI, and what the future holds for India.

 A deep dive into accelerated computing

Dhupar starts off the chat by establishing how NVIDIA has always had a love affair with computing platforms, ever since their company was formed. From its inception, the founders had a strong belief that once someone gets into the business on a first principle basis, it’s always easier to enhance the longevity of the company.

“NVIDIA is a computing platform company that revolves around accelerated computing. With the passage of time, our observation was that most of the world was solving sequential computing problems. With the advent of PCs, a majority of companies were participating in the open supply chain. That’s when it dawned upon us if we will be able to solve parallel computing problems that are profound and need to be performed differently,” he explained.

It is with this very thought that NVIDIA was formed, the first of their use cases being computer graphics. They went on to invent the GPU through a software stack that sat on top of a single architecture. This meant that it was now possible to reach out to game developers and provide them with all the engines that could make games realistic.

“Computer graphics have been very computationally intensive. They have been the driving force of NVIDIA and we continue to work on it. We have always solved parallel computing problems and are now heading towards high performance computing, the third pillar of scientific discoveries. Here, you take the hypotheses or what they call ground truths and then simulate on a gigantic compute to figure out if some of your theories are working or not,” he reiterated.

A transformational era

Over the past year, particularly the last month or so has made it simple for the world to understand how AI is transformational. Whether it’s Chat GPT, generative AI, or similar technologies, these can be leveraged to enhance your productivity.

“Productivity is the main theme of any technology and definitely of accelerated computing. We’ve made observations and we spoke about performance of a different order seven years back at GTC. Today, we are talking about 470 times that performance. Yet, there are several computing problems that need to be applied and solved, and the transformation is very different. When the transformation happens, the entire computing infrastructure has to be rethought and built,” he mentioned.

Speaking about the use cases of accelerated computing in biology and judiciary, Dhupar added that when anything is accelerated, it revolves around an accelerator that is very domain specific. When you create unique libraries, it becomes a body of work that may be difficult to understand. This, in turn, attracts developers.

“When developers write applications on top of that platform, behaviours of companies change, which then brings in a bigger market. In such scenarios, more applications are written,” he explained.

But as we delve deep into parallel computing problems and solve them through redesign, reimagination and re-architecture, there’s also massive power consumption that is taking place. While this challenge wasn’t discussed previously, there’s more chatter about it today.

“If the computing is going to scale at a larger level, we need to be worried about it. How do we save? The promise of accelerated computing is not just speed ups and throughput, but also power saving,” highlighted Dhupar.

With more power constraints, the network plays an important role. Dhupar points out that’s exactly why as a computing platform company, they are so unique in solving those problems.

“This is why you saw open AI speak about it. Even Elon Musk talks about how GPU computing comes to his rescue. Therefore, generally generative AI is about the productivity of humanity. It can be applied to different domains today, and we know how to apply it,” he said.

NVIDIA and the India ecosystem

Over the last four years, India has developed a clear strategy on AI and based it on AI for all. This has also attracted the attention of AI researchers, tech technologists, business people to not just make AI for India in the country, but more importantly, make it the garage of the world.

“India has unique challenges that can be addressed through the application of AI. This can also be applied to the world, because India has a vast population and the size of transactions of tokens is fairly applicable to a large part of the world. The charter of my team is to make a difference and help India accelerate the solutions to some daunting challenges that we have that can be applied to some of the large language models, enabling the entire ecosystem,” shared Dhupar.

That’s the larger purpose of NVIDIA – through the anchors, they are attempting to take the highest end of the workloads and show people how generative AI is quickly able to solve problems.

“There’s a productivity part – this is how we are envisioning its purpose in India. I am very excited we have the opportunity to serve the country,” he concluded.