Silicon is not one of the leading modalities for quantum computers, but it has progressed a lot in the past ~2-3 years. Here are a few key advancements that have happened as of late:
The engineering at those scales is pretty magical isn't it! Getting a whole bunch of individual atoms exactly where they want them.
I wonder what the success rate is - i.e. how many do they build to get one working.
Usually they randomly shoot atoms at the substrate and then just search for a spot (among thousands) where it randomly has the configuration they want. Still pretty amazing.
Can they do that here, they've got quite a few sets of 4/5 atoms which they've interconnected, so that's a lot to get by shotgunning it. I'd assumed they were using something like a STM to nudge the atoms around.
This is a PR release meant to accompany the scientific work shown in the actual source / link. I don’t mean to be argumentative, just, would have taken back the time I spent reading it after reading the Nature version. It’s just “go read Nature” + 3 bullet points + anodyne CXO quotes.
"early days" means that the 1998 computer didn't have qubits that were below the error correction threshold. Now we have hundreds of qubits below threshold. We'll need millions of qubits like these for quantum computing to be useful. If that take decades, this is the "early days" relatively.
Depends on what we mean by "early days on hardware".
If we mean "we've have been working on this for almost 3 decades. That's a very long time to be working on something!". I agree.
If we mean "We just now only have a few logical qubits that outperform their physical counterparts and we'll need thousands of these logical qubits to run anything useful" then we are still in the early days.
What are the real world use cases now, today? The only thing I see in the QC space, are QC stocks and funding paying for the employment of scientific experimentation, which isn't a real world application.
Do I have to wait 15 to 30 years for a series of real world changing breakthroughs that I can already do on a NVIDIA GPU card?
That doesn't exponential at all, in fact that sounds very very bearish.
I think the point being made is that the graphs don't show real world applications progress. Being 99.9999999% or 0.000001% of the way to a useful application could be argued as no progress given the stated metric. Is there a guarantee that these things can and will work given enough time?
- Intel can now do 2D which means a Surface code can be run on these devices: https://arxiv.org/abs/2412.14918
- HRL can now do 2D as well: https://arxiv.org/abs/2502.08861
- They are solving the wiring problem: https://www.nature.com/articles/s41565-023-01491-3
- Their interconnects are high fidelity: https://www.nature.com/articles/s41586-025-09827-w
It's not only early days in hardware, it's early days in practical applications as well: https://arxiv.org/abs/2511.09124
If we mean "we've have been working on this for almost 3 decades. That's a very long time to be working on something!". I agree.
If we mean "We just now only have a few logical qubits that outperform their physical counterparts and we'll need thousands of these logical qubits to run anything useful" then we are still in the early days.
I have not seen any progress or breakthroughs in the QC field at all that are significant.
If the only goal for QC is to try to run Shor's algorithm or to "try to break the bitcoin blockchain" then it is worse than useless.
What are the real world use cases now, today? The only thing I see in the QC space, are QC stocks and funding paying for the employment of scientific experimentation, which isn't a real world application.
Do I have to wait 15 to 30 years for a series of real world changing breakthroughs that I can already do on a NVIDIA GPU card?
That doesn't exponential at all, in fact that sounds very very bearish.
Then invest accordingly, and later reinvest your winnings in a different direction.
There are no real world use cases today. The hardware is not advanced enough yet, but it's improving exponentially.
Quantum theory predicts that they will work given enough time. If they don't work, there is something about physics that we are missing.