Despite all the glamorous promises made about quantum computing, it’s hard to make the argument that today’s quantum computers hold a candle to the sheer practicality of classical computers, especially when qubits need to be cuddled at cryogenic temperatures inside a cryostat. This is worsened by the problem that regular semiconductor transistors do not really appreciate these same cryogenic temperatures, creating an awkward interfacing problem for the controlling electronics.
Now a new pitch here is to create superconducting transistors that will happily work at temperatures near absolute zero. In an article in IEEE Spectrum this start-up – called S-Transistors – and their concept are covered.
By being able to have the control circuits inside the same cryostat, one can forego the absolute mess of wiring that has to penetrate it, and with it one major failure point. Their proposed solution uses the same Josephson junctions (JJs) that are also used for qubits, using a high enough current to briefly make it non-superconducting, inducing a voltage pulse that can be detected.
In addition, JJ-based field effect transistors (JJFETs) are used, with this 2025 paper by [Yusheng Xiong] et al. detailing these structures. Here S–Transistors claims that they are now able to manufacture JJFETs at scale, which would be another major breakthrough that could bring quantum computing just a little bit closer.
They’d be competing with cryogenic CMOS (Cry-CMOS), which can use standard semiconductor production lines to create circuits that can withstand cryogenic temperatures, albeit not quite a the near-zero K level that these superconducting transistors and JJFETs would be capable of.
And the answer is… no. Despite all the fancy promises and being just around the corner for the last 50 years, quantum computers can’t even reach the performance of a GeForce 4MX 440 GPU – which even when overclocked was barley enough to play GTA 3 in 800×600 resolution.
Maybe not good enough for barley, but what about millet?
The highest true reported factored number of a all quantum computer in the world is… 21. All the billions spent have only succeeded factoring 21 in 2026. What a challenge!
Furthermore no quantum computer has ever achieved basic functionality simply because their electronic noise sensitivity continues to decohere qubits before any useful results are obtained. To date ALL quantum computers are failures!
Another no hack post…
And yet you are still reading, so, mission accomplished, I guess.
Hear hear
I do, but my my uBlocked version of this hackaday page is just 975 lines (73 kB) compared to non-uBlocked 1385 lines (100 kB). All those javascripts, tracking and advertising, it just doesn’t load. Damn!
i wonder what happens if you go away for qunatum bullshitery and instead pour all that development into GaAs / InP /any other usufull smecunductor thats not Si.
If we make them superconducting or near super conducting even better, if they have amazing switching speeds perfect.
All the Quantum Computer speed upgrades can be ignored if regular CPU / GPU run at multiple 100 GHz, by default.
No they can’t for this reason. The advertized reason for QC hype is they are computer able to solve exponentially hard problems. You can’t compete with a linear system to an exponential system in the end. But the issue is for exponential performance you need exponentially stable Qbits count that are all a dream at the current time.
Expecting exponential curves to continue un-dampened and un-plateaued is usually a form of madness
Unless of course it’s madness as a function of time that you’re graphing.
“i wonder what happens if you go away for qunatum bullshitery and instead pour all that development into GaAs / InP /any other usufull smecunductor thats not Si.”
Then you get the same number of armchair scientists posting online how none of those technologies had day one performance comparable to existing tech with almost a hundred years of research behind it.
I muse, if qbit computing requires sub-zero temperatures, why not do it in space?
Space isn’t actually cold, it just has poor heat transfer. You can radiate more heat through infrared than you absorb (as long as you’re not in sunlight) but as you get colder you radiate less heat.
It’s really not that hard to get things cold on earth, and you have access to plenty of power.
The issue here is that normal computer chips aren’t designed to work at super low temperatures. Their electrical characteristics change. So if you want to have them close to the bits that have to be cold, then you need to design new ones.
it’s hard for me to imagine that this kind of research won’t have a pay off of some sort or another. but the problem with quantum computing is not the temperatures or the expense, it’s the fact that it’s totally useless. there were salesmen going around 25 years ago saying it would factor large numbers in a single step and nothing remotely similar to that has happened, or is on the horizon. the ‘killer app’ could yet be found, but until it is, finding it is the outstanding problem that prevents adoption or interest.
Thank God Isaac Newton was never concerned about the “killer app”
For a while routes and pathing was on the list. Best plane allocations, delivery routes. All mostly been solved classically.
Maybe protein folding, but AI might just “guess” that one.
Breaking crypto seems to be the only thing left. The way privacy is going, not sure that is in anyone’s interest.