The superconducting-chip claim: change the temperature range, not the process node
Move the logic into the liquid-helium range and replace transistors with superconducting switches — that is Snowcap's bet: tape out RISC-V first, then claim that 28nm-class logic can deliver 5nm-class performance.
The argument · tap a timestamp to hear it
The ceiling on extreme overclocking is cooling, not the silicon
The episode opens out from the overclocking business: everyday chips run on air cooling, higher-power parts move to water, and above that you need sub-zero media like dry ice and liquid nitrogen. Extreme overclockers push voltage to chase the highest frequency they can hit, so the chip runs hotter and hotter and the only answer is an ever more extreme way to pull the heat out. The host says they were an extreme overclocker themselves and once held the number two spot in the world — though it lasted only a day and a half. The reference point they offer is the important part: this kind of overclocking exists to run a benchmark for a few minutes at 8.5GHz, whereas real superconducting computing is not chasing an instantaneous record but logic that comes close to zero power.
In liquid-helium overclocking, the cost is the transfer line, not the helium
Going from liquid nitrogen to liquid helium, physics starts to get in the way. Liquid helium costs roughly 20 times what liquid nitrogen does (nitrogen runs about a dollar a liter), but that is not the main expense: the transfer line that carries helium out of a large storage tank and into a 3-4 kilogram copper overclocking pot costs $5,000 for a single use. Intel and Asus have both done liquid-helium overclocking, and the records have already broken past 9GHz. The host's caution is that material behavior shifts drastically once you get down to these temperatures — some platforms have a ‘cold bug’, where past a certain point the system simply will not start; only platforms without a cold bug are suited to liquid helium.
Superconductivity's appeal is not more speed, it is zero resistance
Overclocking is about squeezing frequency out of the materials you already have. Superconductivity is a different story: electrons flowing through a superconductor meet no resistance, which amounts to moving them for free. Today almost all superconductivity requires extremely low temperatures, and room-temperature superconductivity is still stuck in a research back-and-forth. The inference that matters here: if all the switching logic on a chip could operate in a superconducting state, the energy per operation would be ‘several orders of magnitude’ better than a current laptop. The host deliberately states the claim in full and then pulls it back — superconductivity is still the holy grail, and nobody has it in hand. That sets up Snowcap's entrance.
The former Intel CEO is betting on a temperature range, not a process node
Snowcap Compute has just closed a seed round, and former Intel CEO Pat Gelsinger sits on its board. What they are betting on is being able to build superconducting chips that run at roughly 4-15K — that is, in the liquid-helium range. Structurally these are no longer billions of transistors but a new kind of switch that works at superconducting temperatures: the Josephson junction. At this stage Snowcap's goal is to first produce a RISC-V-class CPU at cryogenic temperature and get its power consumption close to free; the price is that the entire machine has to stay in the liquid-helium range, and the cooling is itself an energy cost.
What is new is that this superconducting logic fits standard EDA flows
Long-term researchers in this field are not scarce: a U.S. defense contractor has built chips integrating tens of millions of Josephson junctions, and IMEC has reached hundreds of millions, with the logic and EDA tooling built out alongside them. Snowcap's point of difference is compatibility: they say this kind of superconducting logic can drop into a standard EDA flow and can be taped out on ordinary CMOS processes, with niobium titanium nitride as the material. The core conclusion the host takes away is that the efficiency advantage superconductivity brings amounts to using 28nm logic to reach performance close to a 5nm chip — note that this is not a smaller node, but the equivalent gain you get from much lower power.
cryo CMOS is only a way station; the real hurdle is scale
Snowcap plans to deliver its first chip in 2026, followed by a RISC-V design that can actually be deployed. For the intermediate state, look at cryo CMOS: it runs at roughly 72-77K, the liquid-nitrogen range, and while there is no superconductivity involved, the low temperature by itself pushes power down. The gap that actually has to be crossed is the one from cryo CMOS into superconductivity, and the host's doubt about it is ‘whether you can get the scale up while keeping high performance’ — getting something running and getting it into volume commercial use are two different things.
Cold boot-up is nothing unusual; keeping a whole machine in liquid helium is
This stretch supplies a real-world footnote on cryogenic tape-outs: the first Zen 1 A0 sample AMD got back, in the period when it was most starved for cash, could only boot at dry-ice sub-zero temperatures because of a design flaw, and was later fixed to run at room temperature. So low temperature during test and bring-up is not remarkable; but holding an entire processor in the liquid-helium range is an infrastructure problem, including the supply of helium — liquid helium is also the lifeblood of the superconducting magnets in MRI machines, and there was a helium shortage a few years ago.
Snowcap is not selling IP, it is selling a closed turnkey system
Snowcap positions itself not as an IP vendor but as a supplier of complete systems. The host's judgment is that precisely because of the infrastructure barrier that the liquid-helium range imposes, the product will end up looking close to a closed turnkey box, and the software stack will have to be built again from scratch. A seed-stage company still has Series A, B and C to get through; that said, they have already brought in advisers and former employees out of Apple, Nvidia and Google. On this subject it is worth staying skeptical, and worth following.
In their own words · checked verbatim
The cost of liquid nitrogen is about, you know, a dollar a litre. And liquid helium is about 20 times that. That's not the main cost with liquid helium. The main cost is actually the tube you need to get it from this massive compressed gas sewer into your massive 3-4 kilogram copper pot. That tube, single use, is five grand.
When you go down to those cold temperatures, the way the materials work is incredibly different. In some chips and in some platforms, you might experience what's called a cold bug. It gets too cold for the system to boot or to work.
There has been a strong effort to make room temperature semiconductors. We're not there yet. Research, if you follow that area, sometimes stalls and sometimes there's a breakthrough. If that happens, one day we may have free energy. Right now, we don't. Right now, superconducting is the holy grail.
But what if you could make a chip that was superconducting? Suddenly, you now have a processor that can do compute for next to no energy. And we're talking orders of magnitude better energy consumption per operation than this laptop, for example.
defense contractors that have designed chips with tens of millions of these Josephson junctions. And then other companies like IMEC doing hundreds of millions of Josephson junctions.
And the headline I got here is that you can essentially build chips with such superior efficiency that you're essentially building five nanometer chips on what is equivalent of 28 nanometer logic.
out comes their first A0 piece of silicon, their first Zen 1 chip. And it would only boot at subzero temperatures using dry ice because of some of the issues that they had in the design of the chip. But they fixed that and it works at room temperature.
Figures
| Liquid nitrogen price | about $1 per liter | 2:01 |
| Liquid helium price relative to liquid nitrogen | about 20x | 2:01 |
| Liquid-helium overclocking record | past 9 GHz | 3:03 |
| Snowcap operating-temperature target | about 4-15 Kelvin (liquid-helium range) | 5:07 |
| Josephson junctions integrated by a U.S. defense contractor | tens of millions | 7:08 |
| Josephson junctions integrated by IMEC | hundreds of millions | 7:08 |
| Snowcap first chip delivery | 2026 (planned) | 8:12 |
| cryo CMOS operating range | about 72-77 Kelvin (liquid-nitrogen range) | 8:12 |
| Boot condition of AMD's first Zen 1 A0 chip | would only boot on dry ice at sub-zero temperatures | 9:12 |
Glossary
- cold bug
- The phenomenon where a chip cooled past a certain point fails to boot or stops working correctly.
- Josephson junction
- A new kind of switching device made from superconducting materials that operates at extremely low temperatures.
- cryo CMOS
- Low-power CMOS that runs in the liquid-nitrogen range (about 77K), short of superconductivity.
How to listen
Hard-tech investors trying to settle whether superconducting computing is a roadmap or a story, and CPU and supercomputer architects working on the chip power wall.
If you already know overclocking and liquid-nitrogen cooling, skip the overclocking setup from 0:00-2:01.