What’s ahead for NVIDIA in life sciences in 2025

Given its exponential and rapid growth, Nvidia has been looking for ways and means of parsing that wealth out into sector-wide investments. Credit: Getty Images / BING-JHEN HONG

2024 has been a good year for US-based technology powerhouse NVIDIA. The company has come a long way from their starting point in the world of video games and computer graphics.

Starting out as a firm primarily focused on graphics cards and video game software, the recent boom in AI and AI speculation has driven Nvidia out of its niche and landed it with a market cap of $3.16trn as a whole. For context, that figure dwarfs one of the largest and most competitive healthcare companies in the scene, Novo Nordisk, which itself vastly eclipses the GDP of its host country Denmark.

Given its exponential and rapid growth, Nvidia has been looking for ways and means of parsing that wealth out into sector-wide investments, in the hopes of maintaining continued growth across the sector as the company’s healthcare arm incubates and invests in numerous start-ups.

Incubating start-ups with plans to spin them out later has been a traditional and long-trusted means of taking immediate capital and sharing it out in the hopes that it will propel both companies forward. Now, the company’s Health and Lifesciences Incubation programme boasts more than 3,500 start-ups under its wing.

Speaking with Medical Device Network Nvidia’s healthcare and life sciences leader for the European, Middle Eastern and African regions, Dr Eva Maria Hempe, broke down how the company is incubating new generations of life science start-ups, and what’s ahead in 2025.

Eva Maria Hempe: In general, we have a huge ecosystem of start-ups. We think about start-ups in two layers. We have the general tier of our inception startup program, which is open to almost everyone. Our criteria for that is that you have to have a website, you have to have at least one developer, you have to be incorporated and you must be older than ten years old as a company. If you tick all of those boxes you can apply.

There is a small screening process where we try to understand whether what you are trying to do fits with what we are doing. Like if you’re doing, I don’t know, landscaping or something like that it may not be a great fit but if it’s something that fits with what we do then we are happy to have them join the Inception Programme, which gets them some benefits on things like training and discounts towards hardware or special programmes.

Then we have what we call the High Touch part, these are the start-ups that we think are particularly interesting and there, we actually also have our developer relations who are working with the developer and looking at the code.

One example is Therapyside, where we really helped them to adopt right-thinking and blueprints to make their architecture better. I just spoke with a colleague of mine who had an interesting conversation with a large pharma company, who was also collaborating with a large range of start-ups. It turned out, that out of those startups, I think 5 out of 6, there are things we can do and help them. Some of them were talking about building learning management systems for AI from scratch.

We tell these start-ups your so resources constrained, you should be using your scarce resources on the things that really differentiate you and not just build something that you can also get off the shelf.

Eva Maria Hempe: In the case of something like Therapyside, it’s about speech recognition. One thing to understand is that we are a platform company, we are providing building blocks, and we are helping people integrate those building blocks in in the best possible way.

Therapyside is a speech recognition model that transcribes the conversations and then we have as a language model, which summarises those sessions into clinical notes. So, it basically helps the therapist to remember things better. Like when a kid’s birthday was or certain titbits about the client so that they can have a better relationship. We are helping with some of the models and helping companies to stitch those together into something much more differentiating on the market.

Eva Maria Hempe: We like to consider healthcare in three buckets, the first one is digital biology. I’m a physicist by training. So, when I was at university, it took a PhD student three years to figure out the 3D structure of a protein, and now with AI we can do this in a few seconds. I can even do it on my phone because now, I can access graphical processors (GPUs) somewhere in the cloud. This is super exciting and will have a huge impact on the drug discovery process. So, companies like Xenia and Insilico have been able to cut down the process of pre-clinical research from five years to 12 months. That is massive.

This is all about falling proteins and creating new molecules, figuring out how some of these proteins and molecules come together. There is one company I spoke to yesterday who want to do like an image of how the pocket in the protein looks like, where the molecule has to dock into and its whole 3D structure. They want to discover a molecule, which fits into the pocket.

That’s interesting because that turns everything on its head in terms of how we are doing things so far.

There’s this whole digital biology space and then we have the digital device space, which is the more mature part of our business. We have been working in this space for at least ten years and built specific libraries and specific platforms for healthcare, but particularly in Europe, we’re very strong with digital devices working with Siemens and GE to kind of design new imaging.

When it comes to start-ups we love all of our children, historically we have done a lot of work in medical devices, a lot of our internal R&D right now is looking a lot at digital biology, but we are really starting to look more and more into the things in between because things like digital health, that is the kind of place where the uses of AI become a lot more tangible quicker.”

Given the company’s rapid growth over the past year, it’s no wonder that it would begin to branch out of its previous technological safe spaces. At the start of 2024, Nividia sat at an approximate stock value of $48 per share in January, by the end of this year the company’s stock value rose to $133. Given that level and speed of growth and the more than 3,500 start-ups under the company’s brand, it’s likely that 2025 will mark another year of expansion for the new healthcare titan.

Speaking on the use of AI in prostate cancer diagnosis, Alberich-Bayarri is optimistic.

“I see a limitless scenario. Models will be able to train on growing datasets which can improve sensitivity and specificity. We are doing a project on a digital twin of the prostate, for example. In years to come, I am sure we will see MRI plus AI in prostate cancer for asymptomatic patients, heading to a preventative scenario.”

GlobalData strategic intelligence analyst Shabnam Pervez says AI is improving diagnostics by “increasing accuracy, speed, and accessibility”, adding that for prostate cancer, the technology can “identify any subtle anomalies that may be missed by the human eye”, which allows earlier detection and treatment plans.

Beyond imaging, AI can also be used to “analyse genetic data to identify risk factors and predict disease progression, enabling personalised treatment plans,” she adds.

And while it may indeed have limitless potential, Pervez said addressing challenges is crucial in ensuring its “responsible and effective implementation in clinical settings”.

“AI in cancer diagnostics faces limitations such as data quality, interpretability, ethical concerns, and regulatory hurdles. Most importantly, it requires a great deal of human trust, diligence, and training, which can take time.”

With more of an emphasis on earlier MRIs, initial costs will increase. However, researchers are already looking at ways to improve cost efficiency. As part of a trial called PRIME involving 555 patients from 22 hospitals in 12 countries, researchers from University College London (UCL) and University College London Hospitals assessed the impact of dropping one of the three stages of the MRI.

The study demonstrated that using the shorter scan allowed specialists to diagnose 29% of prostate cancers – the same proportion as with three-step scans, which reduces cost and makes scans more accessible.

“This will allow more men to benefit from a better, more accurate diagnosis at a lower cost to healthcare systems not only in the UK but worldwide,” said Matthew Hobbs, director of research at Prostate Cancer UK in a statement at the time.

The effects of Sir Chris Hoy revealing his prostate cancer diagnosis have been remarkable – a nearly eight-fold increase in NHS searches for symptoms of the disease is evidence enough. Diagnostics have already advanced rapidly over the past few decades to accompany more men seeking health checkups – whether these tests can overcome barriers to become screening mainstays in healthcare systems is another question entirely.

Australia could be one of the main beneficiaries of this dramatic increase in demand, where private companies and local governments alike are eager to expand the country’s nascent rare earths production. In 2021, Australia produced the fourth-most rare earths in the world. It’s total annual production of 19,958 tonnes remains significantly less than the mammoth 152,407 tonnes produced by China, but a dramatic improvement over the 1,995 tonnes produced domestically in 2011.

The dominance of China in the rare earths space has also encouraged other countries, notably the US, to look further afield for rare earth deposits to diversify their supply of the increasingly vital minerals. With the US eager to ringfence rare earth production within its allies as part of the Inflation Reduction Act, including potentially allowing the Department of Defense to invest in Australian rare earths, there could be an unexpected windfall for Australian rare earths producers.