NVIDIA: AI in the Spotlight

Serge Palaric, Vice President Alliances, OEMs & Embedded Europe at NVIDIA, answers our questions on all things 
artificial intelligence in 2020   

Writer: Tom Wadlow


Graphic Processing Units, or GPUs, have been a gamechanger in the world of computing. 

Driving advances in gaming and graphics, they have also become a key part of modern supercomputers and sparked something of an AI boom. 

NVIDIA is the company behind the GPU, coining the term in 1999 with the release of the GeForce 256. Since then, the American organisation has gone on to create what it calls the brain of computers, robots, and self-driving cars that can perceive and understand the world. 

Serge Palaric is NVIDIA’s Vice President Alliances, OEMs & Embedded Europe. A company veteran of more than 15 years, his role is to ensure that the firm’s network of partners and customers are up to speed with the latest announcements, empowering them with the tools they need to utilise its technologies and supporting them with end customer projects.

“My team and I work closely with our partners, from a range of different industries, to build up an ecosystem that can help companies solve whatever problem they may have with computing and AI solutions,” he elaborates.

Following the unfortunate postponement of Mobile World Congress, Palaric took time out to answer our questions from afar.   


Why did you decide to pursue a career in the tech sector? Have you always had a passion for technology?

Serge Palaric (SP): Technology has always fascinated me. It’s because of my passion for innovation that I pursued a career in this sector. I strongly believe that you will only succeed in your career if you are passionate about it – without this drive, you cannot lead or be disruptive.

The technology sector is extremely fast moving, and a new trend or technology crops up all the time. I really enjoy being part of this movement and having the opportunity to introduce new solutions to the market, power change, and solve complex problems within the wider ecosystem.


How would you sum up your journey with NVIDIA, and what is it about the company that makes it great to work for?

SP:
No two days at NVIDIA are the same and I think that is why it is still an exciting place to work, even after 15 years.

Even though NVIDIA was founded 27 years ago, it still operates very much like a startup. We work at the speed of light and we’re always focussed on developing leading solutions to solve the world’s greatest challenges. 

Every day I wake up and think ‘what problem can NVIDIA help solve today’. Then I set out to make sure that the OEMs and alliances we work with have all the tools they need to take our technologies and start solving these challenges.


I understand one of your specialist areas is AI. What are the key technologies, solutions or trends that are driving AI adoption at present?

SP: AI is powering change across every industry. We’re seeing great adoption of AI for driving innovation across industries including automotive, financial services, retail, manufacturing and edge computing.

Healthcare is one of the key areas where we’re seeing a spike in AI adoption. For medical professionals, it can change the way they work, enable more accurate diagnoses, and improve efficiency. For patients, healthcare innovations lessen suffering, improve care and save lives.

Bringing a drug to market takes, on average, 13 years and $2.6 billion. The rate of success for a drug going through clinical trials is only 12 percent. AI is showing the potential to be a faster, more efficient way to find and develop new drugs. 

For example, MELLODDY – a new drug-discovery consortium – aims to give pharmaceutical partners the ability to leverage the world’s largest collaborative drug compound dataset for AI training, without sacrificing data privacy.

Radiation therapy for cancer patients is a complex workflow that includes modelling the patient, contouring the target and organs at risk, simulating the treatment, planning and delivering the treatment.

One of the most time-consuming tasks in this process is protecting the healthy organs at risk that surround a patient’s tumour and need to be spared from excessive radiation dose. Traditionally, radiation oncologists contour the tumour target volume and organs at risk, deciding how much radiation should be used to treat tumours without damaging neighbouring normal tissue.

To help oncologists develop radiation treatment plans faster, one of our customers, Siemens Healthineers, is using an NVIDIA GPU-based supercomputing infrastructure to develop AI software that enable precision radiation therapy. 


Do you believe the full power or potential of AI is yet to be exploited? What challenges need to be overcome in order to do so?

SP:
We’re just getting started with AI, but to enable us to use it to its full potential there are a number of challenges that must be overcome.

Getting hold of enough good quality data for training algorithms can be difficult. Approaches such as federated learning, which makes it possible for AI algorithms to gain experience from data located at different sites, without sharing it directly, are already helping us overcome this issue.

Finding and retaining fully qualified staff also remains a challenge. But programmes, including NVIDIA’s Deep Learning Institute, are making a difference here by providing online and in person hands-on training in AI and accelerated computing. 


How has the data centre market evolved over the past year or two? Are we still experiencing an exponential growth in the demand for data services?

SP:
The data centre market is continuing to grow, both on premise and in the cloud.

As more and more data are being generated, we’re seeing a growing number of solid AI use cases emerging across different industries and verticals. We’re focusing on providing platforms that support these use cases and verticals, as well as supporting every kind of setup – whether on premise or in the cloud.


Will we get to a point where there is too much data to be stored efficiently and securely?

SP:
No – for AI algorithms, the more data there is to train with, the better. We need to embrace new approaches to managing data and unlocking information that may be hidden within it. Of course, a top priority for a data storage solution is to ensure that it is completely secure.