The word “computer” may be the most expensive positioning error in the quantum industry. Jensen Huang, CEO of Nvidia, recognized it publicly: the bad metaphor slows down adoption, scares investors, and generates absurd expectations.
A word to make them all drop
During the Quantum Day of the GTC, Jensen Huang told a revealing episode. After having declared that quantum computers would still put “10 to 20 years” before being really useful, several companies in the sector saw their valuation fall by 60 %. For what ? Because the market, accustomed to assessing a computer on its ability to replace another computer, expects quantum that it executes Excel faster, or that it “says” a database.
What Jensen Huang points is a Fundamental Framing error : by calling these machines of “computers”, we activate expectations inherited from the PC and the cloud. Result: confusion, skepticism, and pressure on still immature technologies.
It is not a computer, it is an instrument
The alternative proposed: talk about quantum processors, or better, quantum instruments.
“It is not a replacement. It is not made to play Crysis or run a spreadsheet. It is a scientific instrument of precision, a scalpel, not a hammer.” – Jensen Huang
The new analogy is clear: where the CPU generalizes, the QPU specializes. He does not replace, he complete. And its true value is not in raw speed, but in its ability to model the inaccessiblelike electrons mechanics in a material or dynamics of a molecular catalyst.
Nvidia does not build qubits, but structures the ecosystem
Jensen Huang assumes it: Nvidia does not make quantum computers. On the other hand, the company provides the essential abstraction layer to make them useful. It is the same strategy as with the GPUs at the time of Cuda: rather than replacing the CPUs, the GPUs completed them, accelerated, and integrated into mixed stacks.
Three structuring bricks were announced:
- Cuda-q : a language for classic/quantum hybrid systems
- DGX Quantum : full high performance simulation stack for qpu
- Quantum Research Lab in Boston : Partnership with Harvard, MIT, Quera, Quantitinium and Quantum Machines
“We have done the same with autonomous cars. We don’t build cars, but we are all.” – Jensen Huang
Get out of the illusion: lower the bar to just aim
Huang calls quantum actors to change : stop promising universal supremacy and focus on precise use cases, impossible to conventionally solve. What matters is not to beat a CPU on an academic benchmark, but to unlock an industrial problem without alternative: molecular simulation, quantum cryptography, proof of quantum work, design of materials, etc.
The goal is not to be right in ten years, but to remain relevant now.
And for that, it is necessary Change language even before changing the material.
To remember:
Quantum industry does not only suffer from technical limitations. She is braked by A story error. By changing the frame – from an instrument to computer – it could finally get out of the trap of absurd comparisons and enter a useful consolidation phase. Nvidia, once again, does not sell the product, but structure perception. And in an emerging industry, this is often what makes the difference.
