A new paper from the boffins at JPMorganChase suggests quantum computing could soon beat classical solvers-under realistic conditions and with about 73 million physical qubits.
I write for a business audience. So if you’re a tech boffin, do yourself a favour and read the original technical paper. For the rest of us…
If 73 million QuBits sounds a lot, it is. We are nowhere near that yet.
So what just happened?
🔍 This Just Happened
A classical solver is just a regular computer program that solves complex problems using good old-fashioned physics-none of that quantum sci-fi wizardry (yet).
They’re clever, fast, and relentless-like maze-runners systematically testing every path until they find the way out.
They’re also the bread-and-butter tool for quants everywhere, right next to the lunchbox with their name on it and that scary calculator with buttons nobody else understands.
The big news? JPMorgan Chase created a model to show that a hybrid quantum algorithm-QAOA + Amplitude Amplification (AA)-could solve a tough optimisation problem (random 8-SAT) faster than the best classical solver, under specific, realistic conditions.
🎯 Why This Matters for Finance
This gives us a benchmark. Quantum computing won’t just be useful-it may soon be necessary to stay competitive.
If you’re a CEO in finance, the optimisation stack powering your high-value investment products is officially on a timer.
Once the compute arrives, legacy systems will start to feel like Nokia phones in the iPhone era. Customers that can get a better optimised portfolio may be persuaded switch banks for better returns.
Sure, we’re still far from 73 million physical qubits, today’s machines max out around 1,500–2,000.
But the pace of innovation is furious.
Leading firms are talking millions of qubits within 2–3 years.
With an over active imagination, some daisy-chaining, quantum networking, teleportation of entangled data, dots, spins in silicon and hybrid models all slowly coming together, the idea of 70M QuBit Quantum Platform (in whatever format it is delivered) no longer feels completely unobtainable.
If you’re not already playing around with a quantum, you’re behind the curve—and likely a target for disruption.
So What Do You Do While Waiting?
Simple. You simulate.
🚀 Rise of the Simulators
JPMorgan didn’t run this on real hardware. They modelled and simulated what a future fault-tolerant quantum system would need-and proved that quantum advantage is feasible.
If you haven’t gifted your quant team a quantum simulator yet… you need to.
Quantum capability takes years to build.
Finance will be one of the industries most disrupted in the shift to quantum.
Simulators give you:
- A training ground for quantum algorithms
- A sandbox to model and benchmark your use cases
- A no-regret investment in quantum readiness
- The ability to simulate the Google Quantum Supremacy Experiment, on a classical computer with 32 GB of Memory, which is really cool.
And yes-they’re available today. You can run 100s of logical qubits and millions of gate operations on a laptop, using platforms like Quantum Rings.
Think of simulators as the business-class lounge on the way to quantum advantage.
You can get a top of the line Quantum Simulators up and running in under 20 minutes. Tools like Quantum Rings allow teams to run Large-Scale Quantum Circuits Faster Than Ever utilising NVIDIA’s accelerated computing & GPUs. You can even try it for free!
🔄 What the Quantum Algorithm Does
Imagine you’re trapped in a maze. It’s dark.
Your torch is flickering.
You’ve got no idea where the exit is.
Oh-and you’re being chased by a clown with a chainsaw.
Classic nightmare.
Or maybe your last Friday night.
Now imagine you can send out hundreds of clones of yourself to explore every path at once.
One of them texts you the fastest escape route.
You bolt and escape.
The clown shreds your competitors.
That’s quantum computing running QAOA.
🍔 The QAOA Cheeseburger Analogy
You want to build the perfect cheeseburger. You’ve got dozens of options-bacon, blue cheese, ghost pepper mayo.
A classical computer:
- Tries every burger (takes forever)
- Or makes a few educated guesses (faster, but not always best)
QAOA does this:
- Phaser Layer: Scores each burger combo
- Mixer Layer: “Wiggles” the recipe to explore new ideas
- Repeats the process (p = number of layers)
- Measures quantum probabilities to find the most delicious burger
It’s a quantum loop-de-loop that makes the best solutions pop out of the page, and seems slightly magical to the non informed.
🧩 Why It Works (for the Nerds)
QAOA is a variational quantum algorithm:
- It uses parameterised quantum circuits (β, γ)
- Alternates cost and mixer unitaries
- Optimised by a classical algorithm (yes, hybrid)
As depth increases (p = 623 in this case), performance improves-if your hardware can handle it without a hissy fit. (Currently it can’t)
🔧 What’s QAOA Actually Good For?
- Route optimisation (Uber, FedEx)
- Portfolio optimisation (Banks, hedge funds go nuts for this)
- Scheduling (Airlines, factories, cloud computing)
- Machine learning (feature selection, model tuning)
Anywhere classical methods hit combinatorial walls, QAOA has potential to dominate.
⚠️ Is It Perfect?
Not yet.
- Needs error correction (lots)
- Hard to scale
- Sensitive to noise (like most quants)
But when paired with fault-tolerance and smart engineering (as JPMorgan modelled), it can outperform classical machines in the real world—not just in theory.
Think of QAOA as a quantum-powered champion solver for everything your current stack struggles with.
⚔️ The Big Battle: Quantum vs Classical
This paper tackles one of the most important questions in computing:
“Can quantum computers really beat classical ones on meaningful problems?”
In this case: Random 8-SAT with 179 variables. The answer: Yes—but you need deep circuits and big hardware.
✨ The Secret Sauce: QAOA + Amplitude Amplification
AA amplifies the probability of finding a good answer—kind of like turning up the volume on the right solution in a sea of static.
At QAOA depth 623, this hybrid algorithm beats the best classical solver (Sparrow) under real-world constraints.
🧮 The Quantum Budget
To make it work, JPMorgan’s model needed:
- 73.91 million physical qubits
- 1 microsecond code cycle time
- ~15 hours runtime
With better hardware?
You could possibly shrink that to <10 million qubits and solve the same problem in under 3 hours. Quantum systems are already starting to become less temperamental
💼 Why It Matters to Business
Optimisation is everywhere—supply chains, pricing models, logistics, fraud detection, energy systems.
Classical solvers hit walls. Quantum offers a way through.
JPMorgan’s research shows real quantum advantage is achievable—under practical constraints, with costed architecture, and clear technical paths forward.
🤔 Is This Hype?
No, yes, no. Let’s say it is optimistic take.
They didn’t run QAOA on a live machine-but they did model everything that would be needed.
And they did it right.
Quantum simulators like Quantum.Rings now let you:
- Solve real problems with 100s of logical qubits
- Run deep circuits
- Give your team a real playground, not just theory
🔚 Final Thoughts
Previous studies were grim:
“Quantum’s cool, but error correction makes it unworkable.”
This paper?
“Hold my Schrödinger’s cat. Quantum can win. If we play it smart.”
That changes everything-especially for finance CEOs thinking a few years ahead.
🧾 In Summary
- JPMorgan Chase modelled a QAOA + Amplitude Amplification system solving 8-SAT
- It beat classical solvers like Sparrow under realistic conditions
- It required ~73M qubits and deep error correction
- Quantum advantage now has a roadmap
- Simulators are the bridge—get on now or scramble later
📚 Sources:
- Omanakuttan et al., Threshold for Fault-tolerant Quantum Advantage with the Quantum Approximate Optimization Algorithm, arXiv:2504.01897v1
- Boulebnane & Montanaro, QAOA scaling model
