Is it Q-Day?

We monitor the progress of quantum computing to estimate the time when quantum computers will be able to crack encryption. Our information sources include public data from established experts, providers of quantum computing solutions, government resources, and our own proprietary research. It's difficult to identify which source plays the biggest role in picking June 10, 2028, but this date is the primary product of publicly available roadmaps and proprietary research. Our proprietary research builds on observations of the prior 75 years of compute evolution from ENIAC to GPUs, as well as major shifts in personal computers, the internet, and mobile computing. History rhymes.

In 2024 in our internal discussions we estimated that powerful quantum computers would be available after 2027 but before 2030. This was of course a shock to many in the industry who had been estimating 15-20 years, putting the date closer to 2040 than 2030. We are close to the labs and insiders at major quantum technology companies, and we saw that progress was accelerating, that the age-old "always 20 years away" was not something to be trusted. We put more faith in the commercializers than in the academics. This is not to discount academia, but to highlight that commercialization plays a key role in accelerating technology progress. One can see this in the development of the transistor into a viable commercial product.

Now, Google and Scott Aaronson have expressed that 2029 is the year in which Q-Day arrives, which is much closer to our 2028 date. It would not surprise me if their 2029 date moves up — acceleration is not stopping. Interestingly, most of the error correction innovation taking place in the last year has been driven by AI. I suppose one could make the argument that AI is accelerating and not commercialization. Or, perhaps, AI should get equal credit to commercialization. But, I still think the goals of commercialization are much more challenging than the goals to successfully prove a hypothesis (it has to work reliably 24x7 at a customer site in exchange for money vs it has to work once in a lab so the paper can be authoritative).

Why June 10th? I could have picked December 31st and let the wide "sometime before end of the year" carry the weight. But, I felt that there was good reason to believe it could happen before the end of the year, so I picked June, and then for whatever reason my instincts said June 10th. So, I guess the Ouija board had a role, so to speak.

It's possible that we might not ever know the moment when encryption is broken by a quantum computer. There's reason to believe that this will happen and be kept behind closed doors at a government facility, where having such an advantage over adversaries would be a reason to keep it secret. There's another line of thinking which says it's in the public's best interest to broadcast the achievement far and wide so as to alert everyone to the reality that existing infrastructure must be upgraded. I suppose we will know more in time.

The point of elevating the topic of Q-Day is to provide a kind of advanced warning. Now, we don't have the authoritative voice of Scott Aaronson, or the size of audience he has through his various academic appointments and his blog. But, we felt that our estimate was a good one and that society was largely sitting on its hands comfortably believing that Q-Day is at least 10 years away. Our goal is to provide the best info we can about the progress that quantum computing is making, and the impact it will have on society.