QC Ware Runs Enzyme Chemistry on IBM Quantum Chip

On 7 August, a quantum software firm called QC Ware said it had done something that sounds like the future arriving early: it calculated a genuine chemistry problem for a real enzyme using both graphics chips and an actual IBM quantum computer at the same time. Cue the headlines about quantum cracking drug discovery. So let us slow down and read the small print, because the small print is where the honesty lives.

Table of Contents

What actually happened

QC Ware combined its Promethium chemistry platform with IBM Quantum’s 156-qubit Heron superconducting processor to work out the electrostatic interaction energy for nitric oxide reductase. That is a mouthful, so here is the plain version. Nitric oxide reductase is a metalloenzyme, meaning it has metal atoms bolted into its guts, and metal atoms are exactly the sort of thing that make standard chemistry software throw its hands up. Working out how strongly bits of a molecule tug on each other electrically matters enormously for drug discovery, catalysis and materials science, so it is a fair benchmark rather than a toy.

The workflow split the job. GPU-accelerated classical modelling did the bulk of the molecular heavy lifting, and quantum measurements on Heron handled a slice of it. Marry the two and you get a number for a system where the usual approximations tend to fall over.

The caveats QC Ware put in writing

Here is the bit most write-ups skated past. QC Ware itself called this a technology demonstration and said flatly that it is “not currently a fully integrated Promethium product capability.” Translation: you cannot buy this, you cannot run it tomorrow, and the quantum-plus-GPU combo is a proof of concept rather than a button in the software.

Dr Kin-Joe Sham, QC Ware co-founder and COO, framed it carefully too. He said the demo shows classical and quantum computing “can be combined to address meaningful computational chemistry problems,” while stressing that Promethium “delivers high-performance computational chemistry today.” Read that again. The value they are actually selling is the classical GPU speed. The quantum part is the promise, not the product.

Where the real horsepower is

And the classical speed is nothing to sniff at. QC Ware says that for selected workloads Promethium runs demanding calculations up to 20 times faster than the main conventional platforms, turning jobs that used to take weeks into hours. That is a proper number and a proper selling point. The awkward truth for the hype merchants is that the impressive part of this story runs on graphics cards, the same silicon powering your mate’s gaming rig, and the quantum chip is along for the ride as a demonstration of what might matter later.

None of which makes it fake. Bolting a 156-qubit machine onto a serious enzyme calculation and getting a coherent answer out the other end is a real engineering milestone, and doing it on a metalloenzyme where density functional theory usually stumbles is a smart choice of target. It shows the plumbing between classical and quantum systems is starting to hold water.

Why it is worth your attention

The honest read is that this is a signpost, not a destination. Quantum chemistry has been the field’s great white hope for years precisely because molecules are quantum objects, so a quantum computer ought to model them more naturally than any classical box ever will. The catch has always been that today’s machines are too noisy and too small to do it alone. Hybrid workflows like this one are the pragmatic middle path: let GPUs do what they are brilliant at, let the quantum processor chip in where it can, and grow the quantum share as the hardware improves.

So no, IBM and QC Ware did not just cure anything or upend the pharma industry this week. What they did was show the two halves of the machine can talk to each other on a problem that actually counts, and they were refreshingly upfront that it is early. In a sector that loves a breathless press release, that candour is almost the best part.

Did you know: nitric oxide reductase is one of the enzymes bacteria use to help turn nitrogen compounds into gas, which is part of why farm soils steadily leak nitrous oxide, a greenhouse gas roughly 270 times punchier than CO2 per tonne.

Sources

The free stack. One email a week: the AI tools and moves that actually matter, hype filtered out. Subscribe free →