PsiQuantum Raises $1 Billion for Photonic Quantum Computing - September 10, 2025

PsiQuantum just raised $1 billion at a $7 billion valuation from BlackRock, Temasek, and Baillie Gifford. This is significant because institutional investors typically avoid quantum computing due to the technical risks and long timelines.

The Manufacturing Approach That Makes Sense

PsiQuantum uses photonic quantum computing - manipulating light particles instead of electrons in superconducting circuits. The key advantage: their chips work at room temperature using standard silicon manufacturing processes. While IBM needs to cool their quantum processors to 15 millikelvin (colder than outer space), PsiQuantum's approach uses existing semiconductor fabs.

They've partnered with GlobalFoundries to manufacture quantum chips at their New York facility. This manufacturing scalability gives them a potential edge over competitors who need specialized fabrication facilities.

Nvidia Partnership Changes the Game

The Nvidia collaboration focuses on quantum-classical hybrid systems. Most quantum algorithms require classical computers for data preprocessing and postprocessing. Nvidia's CUDA ecosystem already handles these workloads for AI applications.

This hybrid model addresses the practical reality that pure quantum computing won't replace classical computers. Instead, quantum processors would handle specific optimization problems while classical systems manage everything else. Nvidia's betting they can dominate the classical side of quantum workloads.

Scale Timeline and Competition

PsiQuantum targets millions of qubits by 2028 at facilities in Brisbane and Chicago. For comparison, IBM's quantum roadmap aims for 100,000 qubits by 2033. Google's quantum advantage demonstration used 53 qubits.

The quantum computing landscape has seen mixed results. Public companies like IonQ and Rigetti have struggled with commercialization. However, the scale of this funding round suggests institutional confidence in photonic approaches over superconducting alternatives.

What Could Actually Work

Quantum computing shows promise for specific problem categories: molecular simulation for drug discovery, optimization for logistics, and certain machine learning algorithms. The challenge has been building systems with enough stable qubits to handle real-world problem sizes.

PsiQuantum's million-qubit target would enable practical applications that current systems can't handle. Whether photonic quantum computing can achieve this scale reliably remains to be proven, but the manufacturing partnership with GlobalFoundries provides a clearer path than competitors building custom hardware.

Why PsiQuantum's Approach Might Actually Work

Most quantum computers are basically science experiments disguised as products. They need to be cooled to near absolute zero (colder than outer space), require massive infrastructure, and break if you look at them wrong. PsiQuantum's using light particles on silicon chips that work with normal manufacturing - you know, the kind that actually makes products instead of laboratory curiosities.

Using existing semiconductor fabs is genius if they can pull it off. While competitors build custom facilities costing billions and producing dozens of units, PsiQuantum could theoretically scale to millions of qubits using proven fiber-optic and chip manufacturing. That's the difference between a research project and an actual business.

Drug Companies Are Desperate for Anything That Works

Here's the reality: drug discovery is expensive as hell and mostly guesswork. Pharmaceutical companies spend $2.6 billion per successful drug, with 90% of candidates failing during development. BioPharma Dive research shows R&D returns hitting decade-low 1.2%. If quantum computers can actually simulate molecular interactions better than classical computers, drug companies will throw money at it.

Financial firms are equally desperate. Portfolio optimization across thousands of variables is computationally intensive, and quantum algorithms could provide exponential speedups for optimization problems. Goldman Sachs and JPMorgan are already investing heavily. The Brisbane and Chicago facilities aren't research labs - they're designed to sell quantum computing services to companies with actual problems and real budgets.

Cybersecurity is the immediate money maker. Current RSA encryption will be vulnerable to quantum attacks eventually, with NIST estimating cryptographically relevant quantum computers could appear by 2030-2040, so organizations need quantum-safe cryptography and key distribution. Unlike drug discovery promises, this is a known problem with a definite timeline and desperate customers.

Nvidia's Playing the Long Game Perfectly

Nvidia's not just throwing money at PsiQuantum - they're hedging their AI dominance bet. Their Grace CPUs and Hopper GPUs already power most AI workloads, and quantum-classical hybrid systems need both quantum processors and classical AI accelerators.

This is smart positioning. Pure quantum computing is like having a supercar that only works on one specific track. Most real problems need classical preprocessing, quantum calculation, and classical post-processing. Nvidia's ensuring they own the classical parts of the quantum computing stack.

The Institutional Money Says This Time Is Different

BlackRock managing trillions in assets doesn't usually bet pension money on science experiments. Their participation suggests they think quantum computing is transitioning from "20 years away" to "actually happening soon." That's a big shift from previous quantum funding rounds led by venture capitalists gambling on moonshots.

The competitive pressure is real now. IBM's targeting 100,000 qubits by 2033, PsiQuantum wants millions by 2028. Google's quantum supremacy demos are impressive, while IonQ, Rigetti, and Xanadu compete with different architectures. Google's quantum supremacy demos are impressive but operate on different physics with different scaling problems. Someone's going to crack practical quantum computing first, and the winner takes a massive market.

The Trillion-Dollar Question

The $7 billion valuation assumes PsiQuantum succeeds where dozens of quantum startups have failed. They need to:

• Actually build working million-qubit systems
• Solve quantum error correction at scale
• Develop practical algorithms for commercial problems
• Compete against classical computers getting exponentially better

But the potential market is enormous. If quantum computers can solve optimization, cryptography, and simulation problems that classical computers can't handle, the total addressable market hits hundreds of billions annually.

Drug companies spending $2.6 billion per successful drug would pay tens of millions to cut discovery time in half. Major pharmaceutical companies are exploring quantum computing partnerships. Financial firms managing trillions would pay billions for better optimization algorithms. Governments and enterprises will pay whatever it costs for quantum-safe security.

The difference this time: institutional investors with actual due diligence processes are writing massive checks. Either quantum computing is finally real, or BlackRock just burned a billion on physics porn.