GTC Day 3 – Quantum Apology Tour

Jensen was the coroner, not the murderer.

• After single-handedly crashing the valuations of the publicly listed quantum computing companies at CES 2025, Jensen Huang graciously hosted a session at GTC giving the quantum companies the opportunity to prove him wrong.
• The problem is that he was right and all he did in January was point out the obvious fact that quantum computing is very unlikely to be a major economic force for another 15 years at least.
• This was made clear again at Nvidia’s Quantum Day and the publicly listed companies cratered another 10% – 20% as a result.
• Jensen hosted 12 quantum start-ups as well as AWS and Microsoft all of whom were given ample opportunity to make the case for quantum computing.
• The case for it is very clear in that the world needs more and more processing and quantum computing will eventually offer massive increases in processing power at a very low level of cost.
• However, there are many problems:
  • First usability: in that quantum computers are only suited to doing certain types of tasks and when they are asked to do something else, all of the benefits that they offer evaporate leaving behind only the drawbacks.
  • Some use cases are slowly emerging like material science, chemistry and proof of work verification for blockchains, but it is going to be a long time before these become more widespread.
  • Second, technology: where scale, fidelity and reliability all remain massive challenges.
  • Getting more than a few Qubits on a single chip, keeping the error rate down and making sure that the system is reliable remain huge issues and the industry is far from agreement on how these should be solved.
  • Instead, each of the players has an idea of how to solve these issues and is creating a system that is pretty much incompatible with everyone else.
  • This is demonstrative of just how far the system has to go because it is not until there is agreement on the best way forward that the fragmentation can be eliminated and the industry can achieve economies of scale.
  • When I look at quantum computing I see it as being at the same stage that classical computing was in the 1940s and 1950s when everyone was still using vacuum tubes.
• This does not mean that quantum is no good or that it won’t take over from silicon eventually, but it does mean that it is going to take much longer than anyone thinks.
• The real problem here is that there are a number of companies that are publicly listed that have absolutely no business being so.
• In my opinion, it is these companies, their investors and their advisers that are solely responsible for the terrible stock price performance in January as they have gone public with businesses that are not going concerns.
• Take D-Wave for example which was the most vocal in pushing back against Jensen’s January pronouncements.
• This company had revenues of $8.8m in 2024 growing 0% YoY from 2023 from 135 customers.
• With $8.8m, the company lost $42m cash from operations implying that without investing for growth, the company has 4 years before it has to raise more money or go bust.
• Anyone with any sense will immediately see that this is not a going concern as the company is not growing and its real market is more than 15 years away.
• Hence, D-Wave and all of its peers should be financed by private VC money that is well suited to these situations and not be exposed to the savagery of public markets.
• All Jensen did was find a dead body on the street and correctly state the obvious.
• I suspect that in a few years, none of these companies will be public having either gone bust or been purchased by larger companies like Google, Amazon or Microsoft.
• The net result is no change where quantum computing remains by far the leading contender to take over from silicon when it finally runs out of road.
• However, this is not going to happen for such a long time, that RFM could easily have been comfortably installed in retirement for many years before it becomes mainstream.
• This means that Nvidia, with its history of being ridiculously early when it comes to technology, is making the right choice in welcoming the industry to GTC and in opening a research lab in Boston.
• However, it is also right to forget about it for the foreseeable future and leave quantum computing to the boffins while it gets on with the business of monetising the AI boom.