We tell people how atoms move—an interview with Loong Wang

Loong Wang’s first two tech companies were valued at nearly $2 billion when they were sold. He has founded two more since his 2021 pivot to bio-medical technology.

Computing Students Impact Research Videos

We tell people how atoms move—an interview with Loong Wang
We tell people how atoms move—an interview with Loong Wang

Loong Wang might be mistaken for a graduate student with his casual dress and youthful charm. That’s the very mistake we made when we saw him huddled with Canberra-based employees of his latest start-up venture, QDX, which has a dedicated workspace within the School of Computing at The Australian National University (ANU).

Still in his 20s, Mr Wang has built two tech companies valued at nearly $2 billion when they were sold, and founded two more since his 2021 pivot from financial technology to bio-medical technology.

“We tell people how atoms move,” Mr Wang said of his computational chemistry company that applies high-performance computing (HPC) to chemical biology.

We spoke to him about his journey from undergraduate advanced computing to serial entrepreneur.

Taiyang Zhang

ANU CECC: You have founded several companies with Taiyang Zhang. Did you meet in a classroom?

Loong Wang: No, we met just down the road at the Canberra Innovation Centre, where I’d gotten a job with a health tech company.

We got along really, really well and six months on, we said to each other, “Let’s do our own thing.” He dropped out [of ANU Computing] basically straight away. It took me a few more years to be fully convinced. And in hindsight, I think that was the right choice because it allowed me to go deeper into computer science.

I had one of those complicated double degrees that require five years. I probably should have collapsed it all into one and graduated.

ANU CECC: And now you’ve founded QDX with Dr Giuseppe Barca, a Lecturer of High-Performance Computing (HPC). Was he a professor of yours?

Loong Wang: Actually, no. I was a tutor when I was here, and I met Giuseppe through one of my former students who is now a PhD student of his.

I had made a philanthropic donation to help repurpose the old Data61 space, now known as The Hive. [Director of the School of Computing] Tony Hosking and I envisioned it as a collaboration and networking space, and Tony had organised to have a bunch of researchers from CECC come and present their work.

Giuseppe and Loong

That’s where my former student said, “You need to meet Giuseppe.” He showed me a poster about his research and I just… it was perfectly in line with the stuff we were doing at QDX.

I remember the thought in my head was, “Okay, how do I manage not to scare Giuseppe off?” because I was so immediately interested.

I changed my travel plans so he and I could meet a second time over dinner.

I remember going back to Singapore and bringing in the heads of our different teams and saying, “We need to make this the core piece of our technology stack. It will be the differentiator that we’ve been looking for.”

And then I spent the next six months getting to know Giuseppe, doing a feasibility study with him just to make sure the technology really worked. We found out that we had a lot in common—a scary amount in fact, right down to our favourite video game growing up—and we brought him on board as a co-founder.

ANU CECC: What was the video game?

Loong Wang: The original Baldur’s Gate, and we both liked the same character as well: Minsc.

ANU CECC: So, it was a smart move to donate funds to create the space where you met Giuseppe. What’s your philosophy there? Do you think of it as an investment when you donate to create programs like the Computational Biology Talent Accelerator and the Student Entrepreneurship Program?

Loong Wang: I guess, in a loose sense. Early on, we just wanted to get a bit of “street cred” at the ANU so that we could continue to recruit talent.

Many of the best hires I’ve made have been out of the ANU, so, there’s an indirect benefit when I encourage its innovation ecosystem to grow. But more generally, I believe in the growth that can happen in our economy from start-ups and technology. I want to say to students that start-ups are a real path, especially if you’re doing something in deep tech, it’s the path to bringing it into the real world.

This was part of the vision that Giuseppe and I discussed in our first chat.

I’m always disappointed when I see really great computer science talent get scooped up by finance companies in Sydney and they go into optimising spreadsheets. I want to create a cooler funnel, you know?

I want to create an elite deep tech company that recruits the best of the HPC students, the best of the AI students. So, they no longer go into high frequency trading. They go into science and developing medicine and making the world a better place.

ANU CECC: That first company you founded was part of that finance world, wasn’t it? Are these new companies a sort of recalibration for you personally, in terms of your values?

Loong Wang: That was certainly one of the things that factored into our decision to sell our first company. When we started it, we were very much focused on the technology and there wasn’t much understanding about the direction that technology was going. And it ultimately ended up being very much centred around financial technologies. And finance just didn’t really interest me that deeply.

It was intellectually stimulating, but it didn’t have the kind of impact that I was really wanting to have. And so, we decided we wanted to go into life sciences, trying to take our deep tech skills into an industry that was going to have a more profound impact on people’s lives.

We discovered that there’s a lot of technology in life science, but a lot of it is being built by biologists and chemists. Very little had been built by computer scientists.

It’s more common to see life scientists try to pick up computer science instead of the other way around, because in some sense, it’s too easy to get started with programming.

If a biologist learns just the first 5%, they can suddenly whip some scripts together and then “good enough” becomes the enemy of good.

So, we felt that this process could benefit from a talent stream of computer scientists working with life scientists from the get-go.

ANU CECC: It’s interesting that you chose to house the Computational Biology Talent Accelerator with the John Curtin School of Medical Research (JCSMR) instead of CECC, where you studied.

Loong Wang

Loong Wang: Technically, it can be a CECC scholarship. You can be supervised by an academic at CECC. The reason for it is that we are massively strong believers in what the industry calls “closing the loop”. You need to be able to prove it in the lab, that’s the end point for us.

If we had the accelerator at CECC, where there aren’t medical labs, there’s a higher risk that it becomes wholly theoretical. The access barrier for the wet lab is higher. So, it made sense to put it in JCSMR and then gain access to computing capability via CECC. It’s the best way to create a holistic cross-disciplinary group.

ANU CECC: We’ve written about Giuseppe’s work in the past. He keeps on breaking world records for computational chemistry. Using the world’s fastest supercomputers, he can predict chemical reactions at the molecular level. So, the process of developing new drugs and new materials goes much faster because the computer simulations point you in the right direction. Is that what was on the poster that you saw on that night?

Loong Wang: Pretty much. The capabilities that you just mentioned are a subset of the more general capability of what he was developing. So, he developed this really crazy technology that could solve his equations, just insane speeds and really massive systems.

Importantly, he was the first to cross a threshold that hadn’t hasn’t been crossed previously in terms of how many factors you can include. It was enough atoms to be interesting biologically and enough accuracy to be relevant biologically.

And when we met, I said, “We want to put full, proper industry funding behind this. We want to really take it out of the academic setting and explore how to bring this to the real world and actually make a difference.”

It’s gotten even faster and even more scalable in the last 18 months. And it’s also been specifically geared towards the nuances of drug discovery.

But the really nice thing about quantum chemistry is that everything obeys the laws of quantum mechanics. So, you can model anything you like.

We’re hyper focussed on drug discovery at this stage. But we made the intentional decision not to name QDX something specific to drug discovery because there are no limits to where the technology ultimately goes.

ANU CECC: Where else might it go?

Loong Wang: Into material science, so you can design different materials. Into semiconductor design. When you’re trying to design faster transistors, you’re trying to get them closer together. And the problem is that the speed limit that we reached, so to speak, was such that we couldn’t model how the electrons were going to jump around and how they actually start dispersing. The quantum mechanical effects of the electron correlations were too hard to model at those distances. But we can model that now.

The chemical industry as well, chemical reactions and chemical synthesis. You can look at fuels, hydrogen energy development processes, hydrogen transport processes.

Something like 7 per cent of the world’s energy goes into producing ammonium nitrate. But if you can find a better chemical process, you could drop the world’s energy usage by 6%.

But I’m more excited about the medical discoveries.

ANU CECC: Does QDX have any customers yet?

Loong Wang: We’ve got some mid-size pharma companies in the US that we’re working with, biotech startups and research institutions. We’ve got a commercial collaboration with JCSMR. We’ve been working commercial deals since the second quarter this year.

ANU CECC: What does QDX do for its clients?

Loong Wang: The way that I like to explain it is: We tell people how atoms move.

And as it turns out, that’s a really, really interesting thing, if you can do it for enough atoms all at once.

Our clients generally ask us to focus on proteins that are likely drivers of disease, but, to-date, they have never been drugged.

People come to us and they say, “The traditional tools haven’t worked for the last 10, 20 years. We need something else. We need quantum chemistry, and we need it really, really fast and at really large scales.”

And so we sit down with those companies and we help them figure out whether their drugs are likely to bind, where they’re likely to have the effect on the protein that they’re hoping to have, how that protein is going to move and react to the presence of the drug.

Loong Wang

Ultimately it comes down to modelling stuff on computer systems so that you don’t need to model as much in reality, because modelling something in reality is a lot more expensive and actually less accurate than a computer because you have to deal with room temperature and air pressure and all these other external factors.

But it also allows you to iterate much faster, because to design a compound takes weeks and then you have to test it and it’s an expensive and lengthy process.

And let’s say you have a big boom in demand. You could quickly build a big lab space, but then you can’t un-build that lab space a month later when you no longer need it. With computing, it’s much faster and cheaper to rent a massive computer for a week and then you’re done.

ANU CECC: Amazing.

Loong Wang: Yeah, so the economics make a lot of sense. When we partner up with companies we say to them, “Hey, instead of taking the traditional approach which is going to cost you millions of dollars and take many, many years, can we reduce that on both dimensions. We can make it faster, make it cheaper.”

The second thing that’s really interesting to us is not just about doing something faster and cheaper, but doing novel chemistry—chemistry that hasn’t been done before.

Exploring new chemistry in the wet lab is an insanely long process. It takes decades and sometimes wins Nobel Prizes. Instead, we can experiment by modelling on a computer system, and we might discover a new functional group that has never been explored before in this context.

ANU CECC: Going back to Giuseppe’s work, is that what made all of this possible? Or, were you already doing it and his work made it more effective?

Loong Wang: We were already doing it. But, when I first came to computer modelling, my dream was to create a quantum mechanical simulation of everything. I spent a week working on it and said, “Well, that’s not going to happen. That kind of technology is ten years away.”

So we went down a different path. We had a cool framework for a lot of the fundamental technology it required.

And when I saw Giuseppe’s work, it resonated with me so much because, not only was it quantum mechanics, but it was going back to the core physics, which I loved.

It was leaning heavily into physics-based strategies as opposed to artificial intelligence based strategies. So when I met Giuseppe it completely validated how I had been thinking about this industry. There had been quantum chemists building software, but they weren’t high performance computing engineers.

What was needed was a really qualified, high performance computing engineer who knows enough quantum chemistry to build it right.

So when I saw him doing that, it completely aligned with what I knew the industry needed, completely aligned with the direction we had been heading in terms of physics-based simulation.

ANU CECC: So it was your ten-year pipe dream already being realised?

Loong Wang: Yeah, pretty much.

arrow-left bars search caret-down plus minus arrow-right times