Quantum Computing with @hellodavidryan: TDI 24

Threads Dev Interviews

I am finding developers on Threads and interviewing them, right on Threads.

See All the Interviews

Note: The views in these interviews are personal views and do not represent the interviewee’s employer.


“Within ten years there will be everything from small form factor QPUs doing 10 to 30 qubits as a part of roaming networks of devices, through to crazy large quantum devices doing thousands of qubits”

David (@hellodavidryan) on Threads

Today we have @hellodavidryan. How did you end up working in the Quantum Computing space?

My entry into quantum computing came via a headhunter. I’d known him when I was at Red Hat in Australia, working on products such as JBoss, and then kept in touch when I moved to SF and founded a SaaS company.

Quantum Brilliance needed help leading a software release, and I had the specific background of enterprise and devrel, but also SaaS and scaling a product as a founder, which was essential to their mix of otherwise academic-heavy science talent.

I joined as their first Product Manager, immediately open sourced and released the Qristal SDK, and was promoted to Head of Product.

I’ve spent most of the year in airports and hotels and offices, meeting customers and government and industry, and attempting to best find the path of “science to technology to engineering to product”. That’s the challenge in a new industry, and it’s still somewhere between science and technology. Very early days.

With the tech layoffs occurring earlier this year, I’ve been thinking about ways to assist experienced and emerging tech talent to transition into the quantum computing field. Partly because we need all the talent we can get, and partly because this is a ten to twenty year technology arc ahead of us, and those who lived through the birth of cloud computing know how quickly that talent shortage looms. And right now is an excellent time to get involved.

For those curious about how to get into a quantum computing role, I spoke about this at Open Source Summit in Vancouver recently… and there’s a link at the end to a little group we’re putting together to help.

What is the biggest challenge in Quantum Computing adoption currently?

The greatest challenge is one shared by everyone involved, be they quantum companies, businesses, governments, investors, employees or open source community members:

“What does this solve?”

There’s a lot of ways to express this (innovation curves, PMF, etc) but in simple terms it’s just so early. I think of this as “science to technology to engineering to product”. And we are still in the science to tech phase, not just battling incredibly challenging lab work, but finding things we can solve.

This problem can be witnessed in the business strategies of QC companies. IBM sells consulting, Google keeps it in-house, etc. Those vendors are safe due to established revenue elsewhere, but they are under pressure to publish, and you can set your watch to press releases (and generally well meaning if muddled media interpretation) at various cycles relating to public company reporting.

Others like Rigetti and IonQ went public early (the former risking delisting and the latter a huge SPAC).

It’s essential to understand the capital markets and where funding comes from for Deep Tech (or “Frontier Tech” as I like to call it). As this dictates product release cycles, PR and marketing, hiring, etc.

Majority of quantum companies are a mix of VC funded, with significant “sovereign investment” from governments needing to back some players, and a steady zigzag of grants. The only real sales occurring tend to be selling consulting on what’s coming, or selling prototypes to research labs.

That maybe paints a less than flattering picture for devs wanting to just jump in and code awesome quantum stuff. You can absolutely get stuck in without this overhead, but knowing the people like me are tackling these variables is useful, and echoes what we’ve seen recently with early AI/ML teams.

And all of this gives context to why open source is the greatest path into QC. You can engage, learn, and connect, without needing any of the commercial stress that keeps product teams up at night 😂

What hardware is being used for quantum computing?

What do you think of when you hear “quantum computer”? Usually a giant golden science fiction thing, right?

What’s important to know is that’s just one method (and a result of years of awesome work by IBM’s outreach teams popularizing their take on it).

In reality there’s many methods, and some are as small as a handful of rack units. Here’s me next to one we installed at Pawsey in Australia. It’s room temp and virtually plug and play.

It might surprise people to learn that there are room temperature quantum computers, that they can be just a few RU in size, and that they are working in supercomputing centres now. This is true. But these devices are just like the giant ones IBM and others are building in the sense that’s it’s still, as I said in my other reply, battling through the “science to technology to engineering to product” cycle.

What’s exciting is that we’re all trying to find different applications across the full range of ways to create and use the underlying power of quantum acceleration. At Quantum Brilliance that’s a goal of miniaturized QPUs (like GPUs and CPUs) that can be deployed at scale. Think of autonomous vehicle fleets, supply chain and enterprise.

Here’s a glimpse of this installation at Pawsey, where they run the quantum unit alongside a supercomputer.

It also helps to think of quantum computing technology as quantum acceleration rather than a new type of computer. I like to think of it as a Quantum Processing Unit (QPU) because it’s a lot like the GPU revolution.

We can’t just drop a workload developed for CPUs onto GPUs and “it’s faster”. It takes a lot of effort to redesign that workload. And you still need the classical computing CPU. It’s the same for QPUs. They extend rather than replace what we have.

And pro tip: this is why Nvidia is deeply involved in quantum computing. Not just because we use their excellent GPUs for simulation and modeling of quantum computing, but because they are partnering with major vendors to ensure these kinds of hybrid systems are built with the most powerful and collaborative configurations. They’re also an astoundingly smart team over there (I’d personally aim for Nvidia rather than IBM if I were a dev wanting to build a quantum career).

If you take anything away from this conversation today it’s that quantum devices don’t need to be the giant supercooled science fiction mystery machines you see in the media. There’s all kinds being built for purpose, and while the work that IBM and Google are doing at the extreme end of qubit count is amazing, it doesn’t represent the full range being built.

Within ten years there will be everything from small form factor QPUs doing 10 to 30 qubits as a part of roaming networks of devices, through to crazy large quantum devices doing thousands of qubits. What matters more than the pure science is whether they are useful at solving a problem, and whether we can work out how to turn that science into technology prototypes, and then solve the supply chain and manufacturing to engineer that into products. Which is why we need all skills now.

So quantum computers do not use traditional computer CPUs with transistors? Do they?

It’s helpful to think of “a computer” as a general term for a device built to solve a certain purpose. An iPhone, a Linux server rack, a Commodore64, etc.

We can think of “a quantum computer” as a general term for something using a quantum processing unit (QPU) and quantum algorithms to solve challenges that it’s particularly useful for.

That’s going to use the various existing computing technology in and around the experimentation of this new capability. QPUs, GPUs, CPUs, you name it.

At the risk of a boring answer, it’s also useful to embrace the chaos of the early evolution of a scientific concept into technology prototypes and the earliest hint of engineering and product ideas.

There’s always two camps:

  1. Prescriptivists like strict definitions and defend the conceptual purity.
  2. Descriptivists will embrace adaptation and seek to solve problems.

The wave of devs and product teams entering quantum computing is the latter. The timescale of deep tech is always so 😉

I’d urge anyone curious about a career in quantum computing to read @drchrisferrie’s perfectly titled (and hilarious) book. It’s also why I say in my Open Source Summit talk (linked earlier) that “quantum computing is boring” and talked more about utility rather than the science. It’s been decades of stories about cats, slits, and spookies, when we need people to focus on “but what can I really use this for”.

If that seems bold, consider that the only people in tech who care about transistors are those designing chips. QPUs and quantum computing roles still demand a level of knowledge (gates, algos, etc) but it’s not unreasonable to expect a similar abstraction like CPU/GPUs enjoy. I feel we need experienced designers, devs, execs, etc from outside of quantum. A lot of QC companies fail as they are a few scientists with some tired IBM execs burning VC money while bickering internally. That’s a risk.

What can you say about the book you are writing?

The book I’m writing is an extension of my work on open source software in quantum computing. It’s something I needed to know as a product leader but didn’t exist. So I’m writing it.

In short: quantum computing as a commercial product can’t exist without open source software.

And yet: few orgs truly understand the history and tactical use of open source. Not just licenses and community but supply chain, go-to-market, strategies and the specific levels of open source across all of quantum.

It’s easy to feel that we aren’t qualified to create resources such as this. There’s an abundance of brilliant PhDs churning out every year. And yet the real world experience to provide a strategic lens, and advice for action, is as important as a survey.

I’m drawing from experience as a part of Red Hat’s growth to become a $3B+ revenue company, it’s founder as my investor in my own SaaS company, and recent years as a product leader in 10+ countries. So much to learn but keen to share as we do.

At what stage is quantum computing? Is it currently happening or is it all theoretical still?

Hard facts: theoretical with some fundamentally difficult issues around whether anything useful can be done with the current “NISQ era” technology, or if the benefit comes with achieving “quantum supremacy” levels of qubits and… it gets murky and contentious for outsiders trying to make sense of it.

My personal view is that there’s an investment of time and funding and talent to explore if we can get through all of this cycle:

  1. Science
  2. Technology
  3. Engineering
  4. Product

I believe we’re only early in the second stage, and while there are products in market, they are shiny prototypes sold to research labs or supercomputing centers who need to be ahead of the curve.

The phrase “quantum utility” will be more in use in 2024 than previous year’s talk of “quantum advantage” and “quantum supremacy”. This is both a step down in hype and a reflection of the capital markets. Eg VC money drying up, Rigetti risks delisting, and enterprises asking “what can it do yet?”

Someone coming into the industry should be open but sceptical, as PR is a function of appeasing stakeholders or reporting cycles in public markets, and many pre-print papers are “enthusiastic”.

I say this with love and as a veteran of Red Hat where we built a $3B revenue business from what was “a funny hobby” (Linux). I mention this as I can attest what REALLY moves the dial in emerging products is never what you see in public discourse. You’ve really got to get inside the orgs breaking ground.

If I had to pick some quantum computing related companies to study:

  • Q-CTRL has the most amazing product team actually treating product like it matters (and the founder is a force of nature)
  • Nvidia are touching as much of the industry as possible and I’m in awe of the team I’ve worked with there
  • Amazon is doing what Amazon does best (and terrifying the industry)
  • Qbraid is a young and hungry team building a platform from community up
  • IonQ milked the SPAC era and are taking risks

PS: as part of my research on the use of open source in quantum computing I’ve studied and profiled literal hundreds of these companies so drop me a line if anyone has specific questions 📖👍


See the Full Interview on Threads: @ryan.swanstrom • Threads Dev Interview #24 I am finding developers on Threads and interviewing them, right here on… • Threads


Originally Posted

in

by

Last Modified:

Comments

Leave a Reply

Discover more from Ryan Swanstrom

Subscribe now to keep reading and get access to the full archive.

Continue reading