Sep 10, 2025 · 1h 57m · sourcery
Raising $2 Billion to Become the SpaceX of Quantum | PsiQuantum's Pete Shadbolt · Sourcery with Molly O'Shea
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of Sourcery, host Molly O'Shea interviews Dr. Pete Shadbolt, Co-Founder and CSO of PsiQuantum, to discuss the company's $2 billion funding trajectory, its radical strategy to build a utility-scale one-million-qubit quantum computer using tier-1 semiconductor foundries, and the transformative commercial applications across chemistry and materials science.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Molly holds 10.4% of the talking time here. How this is scored →
Molly as informed peer 4.6 Guest teaching 5.0 Guest disagreement 2.2 Molly pushing back 2.1
speaking balance: gold is Molly, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 33:09 AI will not accelerate quantum chip design
Pete directly dismisses the popular premise that AI will shorten quantum hardware roadmaps, calling out semiconductor physics conservatism.
Hardest push from Molly ▶ 38:09 Challenging lack of commercial productMolly confronts Pete directly on having raised nearly two billion dollars without a functioning product on the commercial market.
Biggest teaching moment ▶ 42:00 Explaining DARPA's adversarial red teamPete educates the host on how DARPA deployed 50 scientists specifically tasked with attempting to kill quantum approaches to establish true diligence.
Molly holds their own ▶ 1:17:46 Citing UCLA researcher job projectionsMolly demonstrates deep preparation by citing specific research estimates from Dr. Prineha Narang on downstream cryogenic and hardware manufacturing jobs.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Molly as informed peer | Guest teaching | Guest disagreement | Molly pushing back | Why |
|---|---|---|---|---|---|---|
| Episode Trailer and High-Stakes Highlights | 4 | 4 | 2 | 2 | Molly opens the interview by questioning how PsiQuantum will reach a trillion-dollar valuation with no commercial product yet. Pete outlines their massive funding trajectory and government deals in Australia and Chicago. | |
| The One Million Qubit Strategy vs Incrementalism | 4 | 6 | 3 | 2 | Pete explains PsiQuantum's contrarian stance of bypassing intermediate NISQ systems to aim straight for a million error-corrected qubits, using a rocket-engine development analogy. | |
| Addressing Jensen Huang and Semiconductor Manufacturing | 5 | 5 | 3 | 3 | Molly references Jensen Huang's claim that quantum is decades away. Pete details how leveraging standard semiconductor manufacturing allows 10,000x scale-up much faster than conventional consensus implies. | |
| Collaborating with NVIDIA on Photonics and AI | 4 | 5 | 2 | 1 | Pete describes technical collaboration with NVIDIA across optical networking and software, drawing parallels between historical AI winters and the current quantum trajectory. | |
| Sponsor Message: Brex Financial Operating System | 4 | 5 | 2 | 3 | Following a sponsor read, Molly presses on the path to a trillion-dollar valuation. Pete argues frontier science and quantum simulation of chemistry underpin massive economic value. | |
| Positioning Quantum Computing Alongside AI Compute | 5 | 5 | 1 | 2 | Molly brings up nuclear power revival and uranium enrichment. Pete outlines how first-principles quantum calculations will generate clean training data for conventional AI systems. | |
| Sponsor Message: Turing Intelligence Enterprise AI | 4 | 6 | 4 | 2 | Molly asks whether AI will shorten quantum timelines. Pete firmly rejects the premise, explaining that semiconductor tapeouts and cryogenic infrastructure cannot be meaningfully sped up by AI. | |
| Demonstrating Credibility through DARPA and Government Scrutiny | 5 | 6 | 3 | 4 | Molly directly asks why the public should believe in PsiQuantum without a commercial product. Pete points to rigorous adversarial technical diligence conducted by DARPA red teams and Australian chief scientists. | |
| Evaluating Quantum Valuations and Industry Hype | 5 | 6 | 4 | 3 | Molly asks about public market valuations of quantum SPACs. Pete dismisses simplistic comparison charts and emphasizes operational realities like fabs, power, and site logistics. | |
| Business Model and Future Commercial Revenue | 5 | 5 | 3 | 2 | Pete criticizes venture expectations forcing zero-to-one deep-tech companies to fake early SaaS revenue, laying out a future business model based on compute time sales and selective vertical integration. | |
| High-Impact Commercial Applications in Materials and Chemistry | 4 | 5 | 3 | 2 | Pete categorizes applications, indicating high confidence in small molecule chemistry and materials while expressing skepticism about optimization and quantum machine learning hype. | |
| PsiQuantum's Global Footprint and Site Selection Strategy | 5 | 5 | 1 | 2 | Molly prompts a breakdown of PsiQuantum's physical footprint across Brisbane, Chicago, the UK, and Silicon Valley facilities. | |
| Partnerships with GlobalFoundries, SLAC, and Major Mega-Sites | 5 | 5 | 1 | 2 | Pete discusses utilizing cryogenic infrastructure at SLAC across from Sand Hill Road and clearing legacy industrial sites like Chicago's former US Steel facility. | |
| Deep Dive into GlobalFoundries and Tier-1 Semiconductor Manufacturing | 5 | 6 | 2 | 2 | Molly highlights GlobalFoundries' industrial importance. Pete describes the immense scale and angstrom-level precision inside GlobalFoundries' Fab 8 in upstate New York. | |
| Supply Chain Architecture, Job Creation, and Economic Impact | 6 | 4 | 1 | 2 | Molly cites specific industry employment projections from UCLA researcher Prineha Narang. Pete discusses broad economic leverage from sovereign US supply chains. | |
| Geopolitics, Cybersecurity, and Commercialization Timelines | 5 | 5 | 3 | 2 | Molly asks about geopolitical doom narratives and China levers. Pete confirms cryptographic threats are real but notes commercial chemistry breakthroughs will occur at lower qubit thresholds than codebreaking. | |
| Investor Backing, Institutional Capital, and Debunking Quantum Mystique | 4 | 5 | 2 | 1 | Pete explains how institutional investors like BlackRock demystify quantum physics by evaluating it with the same rigor applied to conventional semiconductor physics. | |
| Molecular Beam Epitaxy (MBE) and Barium Titanate Breakthroughs | 4 | 6 | 2 | 1 | Pete details their proprietary molecular beam epitaxy (MBE) tool growing barium titanate on 300mm silicon wafers, achieving state-of-the-art electro-optic modulation. | |
| Logistics, IP Protection, and Execution Speed | 4 | 5 | 2 | 2 | Pete humorously confirms they FedEx fragile semiconductor wafers across the country and argues execution velocity provides superior defensibility over patent filings alone. | |
| Maintaining Sanity, Immense Responsibility, and Jean-Pierre Luminet's Black Hole | 4 | 5 | 1 | 2 | Molly checks in on Pete's personal sanity. Pete reflects on Luminet's punch-card black hole simulation and the immense historical capital and human effort dedicated to quantum computing. | |
| Capital Efficiency, Thrifty Culture, and Hardware DNA | 4 | 4 | 1 | 2 | Pete discusses maintaining a scrappy 1980s hardware engineering ethos, sharing an anecdote about a facility visitor asking whether to lick a barium titanate wafer. | |
| The 2027 Timeline for the First Useful Quantum Computer | 5 | 5 | 3 | 3 | Molly presses Pete on the exact date for a useful quantum computer. Pete stands by the target of having the Brisbane facility operational by the end of 2027. | |
| Defining 'Useful', Jensen Huang's 'Quantum Instrument' Concept, and Public Messaging | 6 | 5 | 3 | 3 | Molly quotes prediction markets and Jensen Huang's suggestion to rebrand quantum computers as 'quantum instruments'. Pete agrees with the instrument framing and discusses responsible public messaging. | |
| Looking Ahead, Laboratory Energy, and Wrapping Up | 4 | 3 | 1 | 1 | Pete reflects on breaking ground on mega-sites in 2025 and drawing energy from engineering teams on the lab floor before wrapping up the interview. |
Statements from this episode (47)
Shadbolt: PsiQuantum secured $1B AUD Australian deal for Brisbane quantum computer
“Last year we announced these major deals with the Australian government. So that's about a billion dollar Australian deal for us to build a very large scale quantum computer in Brisbane.”
Shadbolt: PsiQuantum is anchor tenant of $500M Chicago quantum campus
“And then also that we would be the anchor tenant Of a new five hundred million dollar quantum computing campus that's being built in Chicago.”
Shadbolt: BlackRock, Baillie Gifford, and Temasek invested in PsiQuantum
“We've had, so BlackRock, Bailey Gifford, Tamasek are all involved in the rounds, and then we have a bunch of new investors also coming into the rounds. But a lot of familiar names from Silicon Valley, and then also some maybe slightly surprising late stage cro…”
Shadbolt: PsiQuantum has raised more private capital than any quantum peer
“That's right to my knowledge, yeah.”
Shadbolt: Useful quantum computing requires roughly one million qubits
“The general consensus is that you need more like a million qubits to do anything commercially impactful with a quantum computer.”
Shadbolt: PsiQuantum allocates zero resources toward small intermediate demo processors
“We've invested in the semiconductor manufacturing, in the networking, in the cooling, in all of the technology that we need to deliver a million qubit machine. And we haven't spent any of our resources On like taking a 20 qubit processor, sticking it online, l…”
Shadbolt: Jensen Huang's 15-to-30-year quantum timeline is likely accurate
“And I think he's probably right on average on that timeline.”
Shadbolt: PsiQuantum's million-qubit bet relies entirely on existing semiconductor fabs
“The only reason that we could afford to make that really extreme bet that we're going to go straight to a million qubits. The only reason that we felt that was plausible is that our technology fits into the fabs and the OSATs and the contract contract manufact…”
Shadbolt: PsiQuantum is collaborating with NVIDIA across multiple tracks
“We are actually working with NVIDIA on a whole bunch of different tracks.”
Shadbolt: NVIDIA is explicit about not building a quantum computer
“NVIDIA have been pretty explicit that they're not building a quantum computer, but they do see a whole bunch of opportunities to contribute to the ecosystem.”
Shadbolt: Quantum will parallel AI's sudden breakthrough via hardware and architecture
“It was really the convergence of two things to my mind the availability of hardware that was built for another purpose in this case, gaming GPUs, and then also progress on the architecture where I see very strong parallels between what we've done and what has …”
Shadbolt: Quantum computers will solve simulations impossible on classical supercomputers
“Our conventional supercomputers suck at predicting the properties of these microscopic things. And that happens to be exactly the set of problems where a quantum computer is perfectly suited and where we will solve problems that can never be solved by any conv…”
Shadbolt: Manhattan-sized classical supercomputers represent the limit of scaling
“I think a Manhattan-sized supercomputer is like, let's put it this way. It is likely the end of the road. Like, it seems difficult to me that you can indefinitely progress beyond Manhattan sized supercomputers.”
Shadbolt: One-million-qubit quantum computer will outperform all global conventional compute
“A single million qubit quantum computer would outperform The net total of conventional computing on the entire planet. So you take all of the supercomputers on the planet, you take every AWS system, every Azure system, you add them all together, and we trash t…”
Shadbolt: Quantum computers will generate training data for standard GPU models
“With that, we should be able to generate extremely precise data in otherwise unexplored territory of chemical space of material space and use that data to drive to train a conventional machine learning system. I think through that kind of mechanism, we can Pro…”
Shadbolt: Semiconductor industry and PsiQuantum use almost no AI in chip design
“We basically don't use AI at all inside the company to design chips. And I think if you go and look at people who are building millions of chips in industry, there's a very small amount of AI, if any, that's being involved in the actual design of those chips t…”
Shadbolt: AI will not dramatically accelerate PsiQuantum's hardware timelines
“So no, I don't think the AI is going to dramatically reduce our timelines. Currently our timelines are mostly determined as far as I can tell by infrastructure, right? We're breaking ground on these sites in Australia and Chicago. These are roughly a hundred m…”
Shadbolt: Most people will never directly interact with a quantum computer
“Quantum computing is the same, right? Like most people will never program a quantum computer. Most people will never directly interact with a quantum computer.”
Shadbolt: Australia's chief scientist found PsiQuantum a 'country mile ahead'
“And then at the end of that, I don't know, 18 month process she concludes that we're a country mile ahead of anyone else.”
Shadbolt: PsiQuantum is one of two companies in DARPA's final quantum phase
“And they have just graduated to the final phase of our program. They're one of two companies to have reached that point.”
Shadbolt: Quantum computing companies are far more likely to go to zero
“Yeah, I think that there is a serious prospect that these companies do that, but there is a much greater chance that they go to zero is basically my simple way of putting it.”
Shadbolt: Basic qubit physics is a small part of quantum scale
“That stuff is really important, but it's a small piece of the puzzle of building a data center sized machine.”
Shadbolt: Venture capital pressure for early revenue warps quantum startups
“We get coached that there is a certain way to do a venture backed company. And the coaching is very often you got to get some kind of revenue by any means necessary really early. And that's certainly true if you're a SaaS company, right? Like if I'm building a…”
Shadbolt: PsiQuantum will not sell its physical quantum computer hardware systems
“Once that machine comes online, we're not going to sell it. Selling a, you know, a 100,000 square foot system is is dangerous and stupid is probably how I'd put it.”
Shadbolt: PsiQuantum employs a 50-person team partnering with industrial companies
“So we have about a 50 person applications team who engage with car companies, materials companies, energy companies, et cetera to prepare them for the existence of that big machine.”
Shadbolt: Quantum Computing Faces Severe Practical Limits for Optimization
“Our view on optimization is that it's really difficult to do like tremendously impactful optimization with a quantum computer. Conventional computers are already surprisingly good at optimization, and the quantum algorithms that we know of for optimization pro…”
Shadbolt: Claims of Quantum Machine Learning Acceleration Are Premature Hype
“Quantum machine learning is really nice to think about. It sounds awesome in like a press release. It's really early is how I put it. Like the development of the theory is not yet at a stage where you can confidently say quantum computing is going to be awesom…”
Shadbolt: PsiQuantum acquired 140,000-square-foot South Bay facility with clean room
“And so now we've moved to the South Bay. We've got 140,000 square foot of space in this new building. We've got 40,000 square feet of former class 10 clean room space.”
Shadbolt: PsiQuantum uses South Bay site as the Foxconn of quantum computing
“And my shorthand on what that building is doing is that it's the Foxconn of quantum computing. So our whole approach is to leverage existing fabs and contract manufacturers and so on. But that building is doing the exceptions to that rule. So where we have to …”
PsiQuantum Spent Over $100M to Establish Manufacturing at GlobalFoundries
“We've spent over a hundred million dollars over the last eight years or so to establish that manufacturing process at GF.”
Shadbolt: PsiQuantum operates a hardware cabinet at SLAC National Accelerator Laboratory
“So we have some of our cabinet, one of our cabinets right now installed at Slack and up and running and supporting chips.”
Shadbolt: GlobalFoundries is the last remaining US tier-one pure-play foundry
“GF is really the last remaining pure play semiconductor, like tier one semiconductor foundry in the United States. So we offshored huge fractions of our semiconductor industry, as far as like a high volume, 300 millimeter foundry that is making chips for indus…”
Shadbolt: 80% of PsiQuantum's Supply Chain Is Based in the US
“Like 80% of our supply chain is in the United States. We have hundreds of companies in that supply chain here in the US.”
Shadbolt: Quantum codebreaking requires a larger system than commercial utility
“As far as we understand still, despite these new architectural results, algorithmic results, you still need a bigger computer to do code breaking than you do to tackle the commercially useful applications that we talked about earlier.”
Shadbolt: Quantum commercial adoption will precede cryptographic security risks
“You should get like an economic early warning sign essentially where you should see quantum computing companies Doing commercially impactful things, making money, and then eventually scaling towards systems that are big enough to be cryptographically relevant.”
Shadbolt: 14 chip and optical companies sought PsiQuantum's breakthrough material
“When we published that paper in nature, I think it was something like 14 regular optical networking and semiconductor companies rang up global foundries asking how they could get access to that material because it is basically the strongest optical material kn…”
Shadbolt: PsiQuantum alone produces its optical material on 300mm wafers
“We're the only people in the world who make that on a 300 millimeter wafer.”
Shadbolt: Reaching the end of Moore's Law will be painful next decade
“Like we're coming to the end of Moore's law. We cannot make smaller transistors. That's going to get really, really painful in the next decade.”
Shadbolt: Quantum has consumed billions and 100k human-years without useful machines
“Quantum computing has already consumed billions and billions and billions of dollars of capital. It has consumed the entire adult life of tens of thousands of our best scientists and researchers. Some total has got to be nearly a 100,000 years of human life. B…”
Shadbolt: The public never cares about supercomputers, only their output
“People have been building leadership class supercomputers for a very long time, and usually people don't care about them, and that's fine. Like nobody cared prior to GPT-III. Nobody cared that Azure and Microsoft and open AI were building big, ugly computers. …”
Shadbolt: BTO is the strongest electro-optic material known to science
“It's made with the strongest electro optic material known to science.”
Shadbolt: PsiQuantum builds 35-layer superconducting wafers in commercial foundries
“So when I look at those wafers that you've seen in our facility, and I think about the fact that there's 35 layers of stuff on that wafer, that there's a superconducting material inside, and that it was made in a commercial foundry next to regular chips.”
Shadbolt: PsiQuantum plans to enable its Australian quantum site by end of 2027
“The plan of record that the company marches to is that we want to have that site enabled by end of 20, 27 in Australia.”
Shadbolt: $1B in revenue marks the starting point for useful quantum computing
“When you're making a ton of revenue you're probably doing something that, you know, within reason you could call useful and like a billion dollars of revenue would be a nice starting point for a useful machine like this. Yeah, definitely what is not useful is …”
Shadbolt: Hyperscaler leaders understand quantum computing requires 1M qubits
“I've actually been honestly pretty impressed that they know you need about a million qubits. They know you need error correction. They know that it's not for PowerPoint and Excel. They understand, you know, these guys have got 20 other things of extreme high p…”
Shadbolt: PsiQuantum will have a narrow customer interface like ASML or TSMC
“The number of people who directly interface with the products of TSMC, ASML, SpaceX, NVIDIA, like how many people do a tape out of TSMC? How many people run an ASML EUV machine is minuscule number of people. And so there are these giant companies where their c…”