Hacker News Reader

Comment by jtchang | Hacker News Reader

The ramifications of the inflection point we are currently at is mind boggling. I had a hard time explaining this last night but we may very well be witnessing the beginnings of a technological transformation era much like when the p-n junction was invented. From the 1940s standpoint it would be hard to envision all we had today.

- Lossless transport of energy - Batteries that don't take any time to recharge - Faster CPUs. Much faster with no heat to burn your lap.

Can I have my flying car now?

View on HN

None of the things you listed are limited by the conductors in them. The efficiency of high voltage AC power lines is limited by capacitive coupling to ground. Battery charging is limited by the cell chemistry. CPU heat output is limited by the resistance of the semiconductors.

Turns out metals (in particular copper) are already incredibly good conductors.

> Battery charging is limited by the cell chemistry.

Yes, but superconductors don't have that limitation, do they? You just dump current into them.

Are you thinking of supercapacitors?

He's thinking of superconducting magnetic energy storage: https://en.wikipedia.org/wiki/Superconducting_magnetic_energ...

Currently they're only feasible as high quality power sources for fabs and other industrial uses because of the operating costs of cooling the superconductors.

Also used for grid stabilization.

So superconductor supercapacitors

More like superinductors than supercapacitors - the energy is stored in a magnetic field (inductor) and not an electric field (capacitor)

There is a theory that with room temperature superconductors you could use an inductor for practical energy storage.

what are the practicalities of storing enormous magnetic fields and then collapsing them to draw power? that seems...not obviously a good idea.

Well, exactly. While there is a way of getting some of the stored energy as an controlled power, most of the failure modes of such a thing leads to uncontrolled release of the stored energy, manifested by the whole thing just somewhat instantaneously turning into very hot gas.

They have low losses when storing power, but power has to be stored in the magnetic field. They have a limit on acceptable magnetic field strength called the critical field though, above which it stops being a superconductor and bad things™ happen. Current SMES systems have an energy density of about ¹⁄₅₀th of current Li-ion batteries.

You are right about the energy density and limitations. I was more focused on the fast rate of storage and retrieval.

We already have incredible supercapacitors.

https://www.eaton.com/gb/en-gb/catalog/electronic-components...

I have 4 at home and blows people's minds when you melt a wire with it. Incredible rapid current delivery.

They don't store any meaningful amount of electricity (which is why they haven't supplanted Li-Ion batteries for power tools) and more importantly they have a cycle-life - they degrade from usage (source: my wife works for a supercapacitor manufacturer).

You also get losses from practical usage - i.e. no one can build a 3V supercapacitor that has decent endurance (you can totally build one which will work, but you're rating it knowing that every cycle is damaging it).

Why do cycles damage supercapacitors? My understanding is that in batteries this is caused by ions not returning to their original spot in the electrodes, but I thought only the electrons moved in capacitors?

Electrostatic force?

Heat?

Overlooking the cell chemistry issue...

... you cannot just dump unlimited current through a superconductor. Once you exceed the critical current density, your superconductor becomes a regular conductor.

And in any case whatever power source you're using cannot deliver infinite power.

Who said infinite power?

It's implied by instant charging. The power would be the battery capacity divided by 0.

No one means literally zero time to charge. A couple seconds compared to minutes or hours is virtually instant for practical purposes.

Let me keep being pedantic and notice that a typical phone capacity of 10000 Joules divided by 2 seconds is still 5 kW of power, so not a trivial amount. I'd think 30 seconds (333 Watts) is more realistic.

If we can even get the charge times for EVs down to the minutes to take to refill ICE tanks. It will basically remove the range anxiety problem.

Best way to get EV 'refill' time to minutes is battery swapping. This already exists commercially, see Nio in China, and it apparently takes as little as 3 minutes.

Similar speeds for charging are impractical because of the power spikes required even if the battery could take it.

You cannot have instant charging, that's not pedantic, that's the discussion, and in any case there is a practical limit to how fast it can happen for similar reasons.

All the work being done on smart home EV chargers that automatically schedule the right time (controlled by grid) to charge overnight are because even at current speeds this wreaks havoc on the electrical subsystems and grid if everyone plug their EVs in at the same time in the evening...

I personally don't have a problem with the current 30 minute charge. If you drive you should stop for 20 to 30 minutes every couple of hours.

What we need is sufficient infrastructure for ev charging.

That’s pedantic. No one means literal instant charging, just figuratively.

Also the thing will be a really strong magnet while it is storing energy. Even if it doesn't rip itself apart it will stick to stuff amd lose energy.

> CPU heat output is limited by the resistance of the semiconductors.

Can't superconductor help here?

Superconducting semiconductors? I doubt such a thing is possible.

Doubt no more. Its called a Josephson junction. https://en.wikipedia.org/wiki/Josephson_effect

It's not obvious to me how it could. Transistors require a semiconductor.

To my knowledge, computers can be designed without any transistors

They can, but then you're talking about a totally different physical scale of computer. Transistors are useful because we know how to shrink them to a scale of nanometers, in particular we know exactly how to do that with transistors printed with lasers onto silicon chips. We'd have to reboot the CPU manufacturing industry with new base materials/technologies.

It's hyper-specialized tech, so it'd probably take over a decade from now to be seen in useful, everyday technologies.

Only a little, one big limitation in chips is RC wire delays (the distributed capacitance along a wire needs to be charged/discharged for a signal change to pass) - due to edge effects capacitance doesn't scale with the area of the wire - but R does (inversely so wires have gotten slower) - so as chips have got denser wires have not got faster at the same rate, when I first started building chips we mostly only cared about gate capacitances - now RC delays are a big deal - if R went to 0 RC delays would too - things would get faster

I'm surprised anyone is even trying that hard to make faster CPUs anymore, now that they're already so fast. You'd think everything would be done by GPUs or dedicated accelerators by now. Especially when so much of the internet is pretty much just serving video files.

But yet there's still so much that could be accelerated even with current hardware, but isn't.

In the RF world (particularly at mm-wave frequencies), even copper has very non-negligible losses. In most passive circuits conductor loss is the limiting factor of performance. No idea if this material retains the same properties at such frequencies, and is compatible with typical lithography or other fabrication techniques, but it'd be amazing if so.

Particle accelerators sometimes use superconducting niobium resonators at e.g. 1.8K offering a Q of 5e10 at 1.3 GHz. But apparently their surface resistance scales with the square of frequency, unlike the square root in normal copper.

So maybe just just t normal conductors like copper, silver, graphene (assuming the latter can be commercially made to surpass silver in RF surface resistance).

Silver is very good for RF power work, so good that I doubt that anything will be able to supplant it on a cost basis because you only need a tiny little bit of it to plate your copper carrier to get really good results. Obviously not perfect, there is still some resistance but it's hard to compete with because of the combination of properties in terms of ductility, flexibility, resistance to deep corrosion and so on.

Superconducting powerlines would be able to transport DC electricity with 0% losses. To put this in context: you could put solar panels in California, and send every watt of power to Alaska or New York, while losing nothing in the transport.

Yeah at those kinds of scales I get doubtful. At those scales the electrical fields themselves begin to exert inertia on things, to the point that you could cut most power lines and they wouldn't stop catastrophically, they'd continue catastrophically (a problem that only gets worse with superconductors, as happened to CERN when they accidentally vaporised a length of superconductor that had become conductive).

It could definitely open up the road to long distance high capacity power lines, but somewhere along the line reality is going to make things difficult. It's not magic after all, just sufficiently advanced technology.

You could bury them or have very high overhead lines. Being able to have near instantaneous, lossless energy transfer would make it worth it (minus whatever impediment the GOP tries to put in the way to bolster gas/oil, but that’s just political).

Could you spot me a link to that CERN incident or on the topic of superconductors becoming conductive (looooool)? All I got for CERN was reports about gas leaks but I find myself very curious about the way superconductors fail now

Here is their own article on the incident: https://home.cern/news/press-release/cern/cern-releases-anal...

Doesn't sound too spectacular but it took them over a year to fix all the damage.

The superconducting effect of LK99 seems to top out at around 150 mA / cm^2. Maybe more research will bear fruit here, but for now we're not looking at HVDC lines.

To my knowledge, that number from the original researchers and no one has yet independently verified it.

External researchers have not even begun to purify and test the limits of this material.

It's the most strange figure from that whole paper, there is just so little context for it.

> CPU heat output is limited by the resistance of the semiconductors.

This is not true anymore. At leading edge process nodes (i.e., smaller conductor pitch), the bulk resistivity of copper changes dramatically because it is increasingly dominated by ballistic electron scattering on the interfaces and grain boundaries: surface area has decreased as a fraction of volume and the average grain size has shrunk to fit inside the interconnect lines.

Alternative metals like Cobalt and Ruthenium have been proposed and to some extent used in production as an alternative to Cu for the low interconnect layers, but they just arrest the trend -- the fact is that the interconnect resistance is much higher than it used to be.

It's also why a lot of the big players like TSMC and Intel are investing heavily in backside power delivery.

Here is an accessible article on the topic:

https://www.fabricatedknowledge.com/p/backside-power-deliver...

(Also: high-voltage DC power is commonly used for long-range transmission and does not have coupling issues)

Yes RC wire delays have becoming more important than gate capacitance for a couple of decades now - if R goes to 0 RC delays do too

I guess the next limiting factor would be LC delays.

No it pretty much is true, well it's the power draw from switching transistors and from leakage losses.

Resistance in the metal wires is not so much a problem for power as it is for signal propagation delay. That is the biggest problem with R skyrocketing in M0/1.

And backside power delivery is to relieve congestion in the interconnects. That's typically one of the limiting factors in complex logic designs now, transistor densities of under 70% aren't uncommon.

That's what "Removing the power signal and signal line to just a signal line would free up space for more transistors." line is about in the link. Power delivery losses aren't nothing, but they aren't a big fraction of energy in today's CPUs.

Eliminating most wire delay would be a huge benefit though, and could be a pretty big revolution. Not just within the core or on the die. You're still going to have to switch those transistors and burn current on leakage though, so no cold CPUs just yet.

> CPU heat output is limited by the resistance of the semiconductors.

Technically true but not applicable... Super conducting CPUs already exist (at below room temp), and they don't use semiconductors. They don't have an exact transistor equivalent but use digital logic circuits such as Quantum Flux Parametron (QFP) and rapid single flux quantum (RSFQ), which are comprised of Josephson Junctions [0] which are made from the same super conducting material. Relying on quantum effects of a gap between two conductors rather than specific material properties.

So while it is true that CMOS are limited by the resistance of both the conductor interconnects AND the semiconductors. Super conducting circuits only appear to need a conductor... There are other drawbacks, such as larger feature sizes due to the nature of super conducting materials, and QFP and RSFQ might be more complex? but this is potentially offset by huge advantages: zero resistance buys not only power efficiency and minimum heat generation (as required by landauer limit), but also allows circuits to be driven at a higher clock, so far demonstrated to 770GHz for one CPU [1]

[0] https://en.wikipedia.org/wiki/Josephson_effect

[1] https://spectrum.ieee.org/superconductor-logic-goes-lowpower

> so far demonstrated to 770GHz for one CPU

That's not a CPU, it's a simple circuit. From the article, "Simple superconducting logic circuits have been shown to operate at speeds of up to 770 gigahertz."

Exotic semiconductor transistors can switch that fast, record being over 800GHz I think. Regular transistors of the type in your phone can probably do around 200-300GHz. Not to say there could not be revolutionary designs opened up with superconducting wires or switches, just the switching speed itself doesn't necessarily tell us one way or the other.

I doubt that regular cellphone transistor can _switch_ at 200Ghz; their GBW might be in that range, but for reliable switching you need clock frequency at most 1/10 of GBW

They can, as calculated by switching delays of simple circuits like inverters and NANDs (so simpler than a FO4).

Actually it's probably more like 200 than 300, I haven't seen 3nm models but transistors are still getting faster.

Good catch, I misread that. Clock frequency is ultimately dependent on the longest path of propagation, so it does depend on the circuit, or CPU.

I'd be interested to know from any experts whether there is potential for fundamentally higher switching speed in super conducting circuits. As you suggest, they are not necessarily directly indicative of the clock speed of a specific real world circuit or CPU, especially considering super conductors don't even use transistors and need to use different building blocks for equivalent logic... but as a rough indication based on order of magnitude of switching speed would be interesting.

It might be possible to create a new kind of transistor that turns on/off by changing the superconductivity of the conducting material.

> The efficiency of high voltage AC power lines is limited by capacitive coupling to ground.

And to this point, HVDC has been slowly rising as a viable alternative beyond just undersea transmission.

A mistaken belief is that AC is more efficient than DC. What AC is is more easy to transform from one voltage to another (until somewhat recently). That makes it easier to run AC at 1 MV and then step it down to 240V for residential applications.

The cost is as AC voltage goes up, capacitive resistance increases. This is part of the reason why high voltage lines have such huge towers with the lines far apart.

An HVDC line, however, can be put underground (or water) without suffering power losses. It's voltage can go well above that of AC voltages with the only limit being how much insulation we need for the line.

Super conductors are nice, but only in the "now we can run 1 billion amps at 100V" sense to avoid the capactive resistance. Without any sort of special materials you get most of the benefits of super conductors by using HVDC. The only real downside is high voltage DC is still super dangerous. Cut the insulation and you've got something that can arc meters whereas a superconductor at lower voltages would be about the same danger as any other conductor at lower voltages.

Minor nitpick — "now we can run 1 billion amps at 100V" — there is a limit to how much current you can put down a superconductor. For example I think commercial YBCO superconductor tape will do roughly 1000A/mm^2 (and even that requires being significantly below its critical temperature).

That’s a pretty high current density compared to what’s feasible with copper — don’t get me wrong — but a billion amps would still require a pretty huge cable even with a superconductor.

Presumably everything in the chain, every junction, would also need to be a pure super conductor? Anything else would be susceptible to instantly melting or vaporising.

So unless someone comes up with super conducting welding I guess that would mean no junctions.

Not only that, the magnetic field around the wire at 1 bln amps will be way too large. It will like pick up nails from the ground easily.

AC is also easier to switch under load.

"A mistaken belief is that AC is more efficient than DC. What AC is is more easy to transform from one voltage to another (until somewhat recently). That makes it easier to run AC at 1 MV and then step it down to 240V for residential applications."

DC is also used because a conductor can carry more current per kg than AC due to not having any skin loss. You can use smaller wires, less metal, less dollars.

>The efficiency of high voltage AC power lines is limited by capacitive coupling to ground.

As often in engineering, the sweet spot is where multiple factors have ~equal contributions, so even a 50% win in one of the factors can't give you that much benefit.

For high voltage AC power lines it's Ohmic losses, corona discharge losses, and inductive losses.

But it is. A significant part of electronics engineering is dedicated to heat dissipation.

So what do we stand to gain?

At least a slight improvement in efficiency of transmission. And a huge improvement due to simpler heat dissipation needs.

So this might mean way faster battery charging - like 500 miles worth of charge in 5 mins. We’re mainly limited by how hot the batteries get during charging.

We're about halfway there already, just by cooling: 250km in under 4 min (1).

(1) https://www.tue.nl/en/news-and-events/news-overview/13-07-20...

I love that electric vehicle charging can be expressed in speeds (that is, distance divided by time). 250km/4min is 3750km/h, a touch over Mach π.

I like this. Woah, your car charges at 4000 kilometers a hour? Mine only does 3000!

This is incredibly exciting. The biggest hurdle as yet to mass adoption.

We’re not talking about charging existing batteries but superconductors where you just “trap” large amounts of electricity in the superconductor for future use.

Unfortunately magnets put forces on themselves, which means you’re limited to storing up to about as much energy as it would take to break the thing making the field; or, to put it another way, it’s about the same range of energy densities as springs or rubber bands.

Earlier discussion: https://news.ycombinator.com/item?id=36954783

Does that mean you would get equal(but probably insanely more dangerous) performance with a giant miles long spring instead of the proposed superconducting loops?

Do you mean superconductors create super capacitors, or how do you want to trap the charge?

Or store in coils, as "magnetic energy"(forgot the name for it)

> - Lossless transport of energy - Batteries that don't take any time to recharge - Faster CPUs. Much faster with no heat to burn your lap.

Finally someone is speaking English in these room-temperature superconductor threads that I can actually understand and get excited about!

> None of the things you listed are limited by the conductors in them.

Never mind. :(

> Can I have my flying car now?

I ditto the sentiment. But we don't want literal flying cars. Well, self driven flying cars. Humans have enough problems when they're driving on the ground on ground made for driving.

I wonder if self flying cars are easier to make since every object in the air is an obstacle. This is less the true for ground transport since sometimes it may seem like an obstacle but it isn't (e.g. just a marking on the road).

Clouds are visible, but may not be an obstacle. Wind is invisible, but may be an obstacle.

There are far fewer things to run into in the sky.

true.

However the failure cases at those heights and velocities are far worse. There's several orders of magnitude difference between an airplane license and driving license.

> Batteries that don't take any time to recharge

Huh? Is this actually a thing that this enables? I don't initially see how

Resistance is what makes things hot, and heat is what makes dumping huge amounts of charge current into batteries a bad idea. No resistance → no heat → no need to charge with low current†.

Another way to say it is that, with a superconducting wire, you can make the wire as thin as you want and still pass the same amount of current through it, without melting the wire. Picture using a USB-C cable to charge your car.

† (There'd still be a current limit due to the heat generated by the chemical reaction that rebuilds the battery's voltage potential... if said reaction is exothermic. Some battery chemistries are endothermic when charging!)

Wouldn't the battery itself still have resistance? Or is the superconducting material itself a battery?

Depends. A single battery cell would have nontrivial resistance, yes.

But a big bank of batteries, like are in an EV? Very hard to give them enough current to heat them up. Most of the "heat problem" is from the bottlenecked current path into the car; once you fan out across all the individual cells, each individual cell isn't receiving much current.

And a bank of supercapacitors? You could charge it effectively instantly.

The current is limited by what the battery chemistry can take, not by the cables. This is why the first 80% can be charged quite fast in modern EVs, and the last 20% are really slow.

Additionally you need to have the current to deliver in the first place. Having a grid that can dump 25-100 kwh into any given car in a couple of minutes is no small task if everyone is doing it.

The utilization factor would obviously be much lower than it would be if everybody charges at a lower rate so if the total amount of energy is equal that just means that individual vehicles will spend less time charging, and the grid will see - roughly - identical utilization on average but the peaks may be higher.

Probably worth pointing out that the peaks and troughs are what are challenging to deal with. Generators aren't generally great at changing output super fast.

I keep hearing battery tech is getting good, and the research I've seen suggests that more storage on the grid would improve efficiency by a lot, so I don't know if it would even pose a particular challenge if that sort of demand arose.. but overall utilization isn't really the limiting reagent.

Sure, it would make the wiring smaller and more efficient. But I also don't see how it would help in the chemical energy transfer to charge the battery.

What I was trying to say is, with some battery chemistries, the current (heh) limiting step for charging speed is the wiring into, and of, the battery, rather than the safe reaction speed of the battery chemistry itself. We could safely "crank the chemistry" by an order-of-magnitude or more if we could get the desired current into the battery without the wires+electrodes conducting undue amounts of heat into the electrolyte.

No, it’s not. It’s the chemical reaction the limit. In li-ion for example you will create dendrites when charging/discharging too fast or too deep. This is the cause of the relatively short cycle life.

According to the paper, this material stops superconducting at about 150mA per cm^2 of diameter, meaning that a 1cm-thick cable made of this material could conduct up to 150mA before the current is too much and it stops superconducting.

If my math is correct, then for a basic 500mA USB device, that would mean a cable a bit over 3 cm^2 in cross-sectional area, or about 2 cm across (for each of the power and ground leads, at least).

Alternately, a cable of just over 1/2cm in diameter (for power and ground, each) could charge a rechargable Ni-MH AA battery in about 12 hours and 40 minutes.

Tl;DR this is absolutely revolutionary science, if true, but we're definitely Not There Yet.

Superconductors also have an inherent current limit above which they go back to having a resistance.

I’m not an expert on this, but I think superconducting wires have an current limit, as a current flowing creates a magnetic field which the superconductor has to repel. I read that the paper states a very low current limit for LK-99, meaning it loses superconductivity once a very modest amount of current is passed through it.

That is.....kind of a huge limitation of the technology lol. Still very cool but less hype

I believe the implication is that LK-99 is basically a demonstration of an entire class of materials which should have room-temp superconduction properties. IE we can enumerate through the entire class and find the ones with the properties we want.

There is no reason to assume that even if it's real.

In fact, the tight tolerances of this seem to indicate the opposite.

At least according to this wikipedia chart on superconductor discovery timelines[1], it seems like most discoveries aren't one-off.

I have no knowledge in this area though.

[1] https://en.wikipedia.org/wiki/History_of_superconductivity#/...

A limitation...at ambient temperature and pressure.

Usually this is an optimization frontier, where something that has tetchy critical current/field at high temperature is going to have very good critical current/field at the same temperature as a lower-Tc superconductor.

If it superconducts at all at room temp, cooling it down even to 200K (about dry ice temp - quite cheap to do) could get you something very usable.

It's hard to tell what the critical current density of LK-99 is, because their sample is porous and probably very impure. They measured the critical current they could pass through a sample, but the conducting cross-section is somewhat unknown. Its high critical temperature suggests that it should probably have a higher current capacity than other superconductors. That said, in the extremes, current density is also limited by tensile strength, because electromagnetic coils repel themselves.

Couldn’t you bind it in resin?

Is there any indication of what the breakdown voltage of LK-99 might be?

> you can make the wire as thin as you want

No, superconductors have a specific current above which they stop superconducting so you will want to stay away from that limit. This particular superconductor has been presented with a very low Ic (150 mA in the original paper0 which would not make it particularly useful in such applications but future iterations (assuming it is all true) may improve on that (they should otherwise we have the equivalent of a superconducting straw).

I assume this would rule out things like fusion reactors, MRIs, and other high energy stuff. Would it still be revolutionary tech with a 150 mA limit?

Yes, it would be upending just about everything because the race would be on to improve on that. Think of it this way: once you show that something is possible at all there will be substantial funding available to improve on it. As long as you can't show that it is possible at all you're on your own. So if it works and that 150 mA is the limit then you can expect a ton of effort to be expended to improve on that and I fully expect those improvements not to take decades to show up. The more interesting question is if it really is that low of a limit what the reason is for that and I don't recall seeing any explanation so far.

On another note: a superconductor that can only do 150 mA / cm^2 seems intuitively strange, as though that figure is somehow off, it's a gigantic cross section for such a small current. It is very well possible that this is somehow an error in the reporting or an actual measurement on a thin sample with small cross section. So there are many explanations possible and only one of those is a true limit of the material.

The current hypothesis is that most likely whatever they made is not a pure sample of the material which actually superconducts - this is expected, since when you make YBCO superconductors you also tend to get low yields (i.e. ~20%) that actually superconduct.

So it could be the whole sample, or it could one micron-sized link of grains of whatever the "real" material is running through the sample.

> The current hypothesis is that most likely whatever they made is not a pure sample of the material which actually superconducts

That has been the hypothesis from day #1.

There are many things that seem like electrical resistance but are different phenomena. Capacitive reactance, inductance, "radiation resistance", etc. Superconductors don't prevent any of these effects. But, these effects are usually smaller than ordinary resistance.

Heat from power transfer is not the problem with current battery tech. We are already capable of delivering 350kW worth of power into EV batteries. The limiting factor is not the power cable delivering that power.

Thick cable, high voltage (900V typically) and everything is fairly manageable. Assuming we could consistently charge at that 350kW we could fully (0->100%) charge an 80kWh ev battery in 13 minutes. That's not slow.

The limiting factor is the battery chemistry, not the wire chemistry.

It’s a zero resistance wire. Build a loop and pump electrons in. Need them back? Connect an off-ramp.

There are limits how much you can pump into it, it isn’t magic… but it almost is actually.

https://en.m.wikipedia.org/wiki/Superconducting_magnetic_ene...

> Less than 40kJ/L

Pretty sure that includes the volume of the entire cooling system needed to keep it superconducting.

The entire reason the current (possible) discovery may be revolutionary is that it doesn't need a cooling system.

But the current density limit makes that not super effective right?

Nah, the GP is just completely out of reality.

We won't see lossless transmission in a very long time, and no place where an aluminum cable is too expensive today will become viable because something a million times more expensive is 9% more efficient.

Batteries won't see a revolution because of this, there's simply no reason for them to (but they are currently in a revolution, and there are more to come). AC storage in the superconductor will probably be the most expensive storage mechanism you can buy, and flywheels will keep having atrocious energy density, they won't even get twice as good. But it will completely revolutionize some niches in storage.

This won't replace metal layers in CPU for a really long time. Superconductors are hard enough to make, CPUs are absurdly hard to make, and the wins on power savings aren't very large. If people make superconducting chips, it will be ones where the superconductors do active switching, what is much farther away and can enable much faster CPUs too.

I really wish people would stop repeating those. If you are going out of your way for an outlandish claim, I'm much more interested on discussing if this can replace rockets for near Earth space travel than those absurd costly low gain things.

It's interesting you are saying superconductors are hard to make because... if this one really is a superconductor it's pretty easy to make. YBCO is also not particularly hard to make either.

The semiconductor industry is also very happy to do hard things for marginal gains. They've spent a decade swapping out semiconductor junction materials at enormous expense because there's no other option.

The idea that reducing power consumption is also not large enough to matter is...yeah, detached from reality. Thermal density has been an enormous problem with increasing CPU feature densities. CPUs already run hotter then a kitchen hot-plate, which is why so much effort has been put into dynamic throttling and other tricks - you straight up can't run CPU circuit elements full-power for very long without the risk of frying them, or requiring a cooling system which is impractical for widespread deployment.

>This won't replace metal layers in CPU for a really long time. Superconductors are hard enough to make, CPUs are absurdly hard to make, and the wins on power savings aren't very large. If people make superconducting chips, it will be ones where the superconductors do active switching, what is much farther away and can enable much faster CPUs too.

I would've thought the main wins would've been reduced heat generation. Like you said, power savings would be negligible. But at the data center scale I'd imagine that reduced heat would result in reduced air conditioning power costs. And potentially with thermal constraints removed, it would allow for more compact packaging per server as well.

Like you said though, I doubt it would happen any time soon.

> I would've thought the main wins would've been reduced heat generation.

Most of the heat in a data center is coming from power burned in the devices themselves not the power transmission to those devices.

Sort of like how a space heater is hot at the heater portion, not the power cable going into the heater.

You aren't losing more than a W or so to transmission for every kW of power delivered. (in fact, you are generating more heat from the AC->DC transformation)

Typically, power cords and wiring is 15 AWG, which has a 10 milli-ohm/meter resistance and runs as 120V AC (maximum of 15->20A). So, 1000m of power cord running at full load would result in 150->200W of heat from the power cable. Meanwhile the server is generating 1.6->2.2kW of heat. (Assuming a single very long 1kM route is servicing the server).

Cut the cabling distance to a couple of meters and you can see why nobody worries at power consumption at that point.

I meant in the integrated circuits rather then power supply.

You can start to get to the answer if you look why metal layers aren't made with silver. If metal resistivity was that important, nobody would pass the chance of almost doubling the conductivity with a much smaller change than people are proposing here.

The reason they don't is because yeild is the one most important variable on the entire process. And silver adds enough complexity to decrease it. (A few chips do have silver layers. People manage to use them when the process is mature enough and the added complexity gets tamed. Including something like YBCO into a high-performance chips manufacturing process is a half-a-century project; LK-99 can't even be reliably done yet.)

> something a million times more expensive

Lanarkite and copper phosphide aren't expensive, other than the government controls on red phosphorous (because it can be used to make meth).

They way I have described this is like the Wright brothers have just demonstrated the first power flight, and we are already planning the schedule for the A330's from LA to London.

It is exciting stuff but there is a very long way to go if true.

> Can I have my flying car now?

Let's not forget the flying skateboard of the film "Back to the future". I loved it in the film and it's a dream that I still have today - I'm now almost 50 years old so I would probably crash and get killed by using it, but I would still give it a try :)

Unfortunately, while you can indeed build hoverboards with superconductors and they do work, you still need a magnetic surface for it to ride over. I don't believe generalized hoverboards that will work on all surfaces like BttF are possible.

What about Earth's magnetic field?

It's moving all the time. Next year you won't be able to skate the same locations as of today because of this :(

Well not with that attitude

Honestly, with that kind of superconductor, it may be easier and cheaper to cover the ground in superconducting material, and keep the magnets (or superconducting electromagnet) on your hoverboard!

"Hoverboards don't work on water!!"... or any non-magnetic surface :)

Wouldn't a hoverboard with rockets on the bottom and back work? I'm not saying it's feasible, just that it's utilizing the same basic science as jets and space travel.

I think if you look up the smallest rocket that is capable of even lifting you off the ground on an instantaneous basis, let alone over time, you are going to be somewhat surprised at the size. I think you will also decide you don't particularly want to be at point-blank range to this rocket when it goes.

I mean, theoretically, if you loosen the definition of "hoverboard" enough, it might be possible to create something that hovers with you on it, but I don't think you'd accept it as a substitute for the movie hoverboard.

Where is my hoverboard?

https://www.youtube.com/watch?v=eOH15_pqWZ4&ab_channel=SETHS...

There's no need to wait for flying skateboards, you can buy a regular skateboard at any shop around town and crash and get killed by using it.

(Said in jest by someone who recently turned 41 and hurt his back playing video games last year)

I remember the feeling, while skateboarding, of the change of the roughness of the ground - some streets (or at least portions thereof) were very smooth and that felt already quite like flying => I wonder how that would feel with 0 surface roughness :)

>hurt his back playing video games last year

Some extreme force-feedback device? :)

Picking up the CRT and throwing it out the window?

You are obviously joking but still, give foiling (as in pump foiling, wing foiling or eFoil) a try. It's not the same but the closest you can get, at least within 3 feet over a water surface (which makes crashing a lot more benign)

I mean, yeah, you'd need a very generous marketing budget to make a skatepark full of magnets so we can have hoverboards. Like the one Lexus already built: https://www.theverge.com/tldr/2015/6/25/8846785/lexus-hoverb...

Did we read the same paper? This is a ceramic.

To kill your buzz:

Liquid Nitrogen superconductors shocked the world in 1987 but have hardly changed it. They have some applications but we don't have transoceanic power cables, superconducting supercomputers, MAGLEV trains everywhere, etc.

You could have made the case that the cost of liquid helium cooling put traditional superconductors out of reach for most applications, but liquid nitrogen cooling is not difficult at all. Unlike helium, nitrogen is a renewable resource. Cuprate superconductors have been held back by issues that have had nothing to do with cooling and even if the new room temperature superconductors are for real, it's possible they'll turn out quite like the cuprates.

One strange thing about cuprate superconductors is that the theory is not understood despite being a "holy grail" for more than 35 years. It fits the schema of problems like dark matter, neutrino masses, the matter-antimatter and how energy gets coupled from a quasar accretion disk into a jet... Cases where a devilishly hard problem can go unsolved for the working lifetime of a physicist.

Reminds me of a story: I was in college physics in fall '89 and our professor was telling us how he and his son spent the summer in Alaska prospecting for whatever material was all the rage in superconductors at the time. He was explaining that when superconductors broke the liquid nitrogen temperature, it was a game changer. He said "If you buy it by the gallon, liquid nitrogen is cheaper than beer."

To which a student replied "You buy beer by the gallon?"

ever bought a keg?

It might even all come fast enough to save us from climate change

(But I'm not getting my hopes up)

we already know how to stop climate change: actually stop burning shit and deploy existing technologies quickly. the problem is lack of will, not lack of technology.

corollary: anyone trying to say we need fancy new technologies like fusion/superconductores/supercapacitors isn't actually very interested in stopping climate change.

> the problem is lack of will

Lack of will which fossil fuel shitbirds spend billions enfestering, with tobacco company style tactics. They knew exactly what the fuck they were doing for the last fifty+ years.

We probably agree on that, I'd just like to focus the blame where it properly belongs. Plenty of people care a lot about climate change, just as we care about plastic pollution and inequality, and I'm pretty fucking tired of being gaslit about it all.

There are much more effective ways to geo-engineer earths climate. Their problem is the politics that surround all the options. Sooner or later we won't have other options.

Nuclear fusion would be quite handy. That's near limitless electricity from minimal input, with mostly safe failure modes and no nasty emissions.

Decent superconductors might enable, less lossy power transmission across distance, Maglev at scale, or perhaps initially, lower power consuming, small devices - every little helps.

I do agree with you though - that lot ain't any good right now. Don't allow yet more licenses in the North Sea etc ...

> mostly safe failure modes

I believe they call this thing “harnessing the power of the sun”. I’m mildly sceptical about the safety of it’s failure modes.

Not to say I don’t still want it, but…

Fusion is not easy to sustain. The massive pressure in the center of the sun is instrumental in it working there - something we lack.

yes but with all this power we’ll be able to render much more accurate models of our demise

Just to be a bit of a realist do we know if this material is malleable or practical to make intifrates circuits? Is it possible to make large single pieces of it? Don't get me wrong even if the awnser to all of these is no it's still probably the biggest single material science breakthrough since the transistors but we aren't necessarily going to be applicable to all the theoretical applications of semiconductors

> Much faster with no heat to burn your lap.

Computation inherently creates heat, that's not something that superconductors will change.

This is repeated over and over again but that's only a very small fraction of the kind of power that a computer uses. By the time you're talking about reversible computing all the low hanging fruit has been plucked and there are much, much bigger sources of loss. The biggest one impacted by superconductivity if (and that's a really big if) it can be used for the interconnect layers ('metal') in a chip and for the circuit traces outside of the chip that you can cut the charge and discharge time for the gates of the transistors in the chip down to a minimum. This in turn changes the power consumption of the chip because the transistor is either 'on' or 'off' and spends much less time on the transition in between where it is more of a resistor than a switch.

So it isn't determined whether or not it will be changed but it could be.

Flipping a bit from 1 to 0 releases heat (because you can't just drop the 1 onto the negative/ground)

Resistance in non-super- conductors wastes electricity as heat.

From "Thermodynamics of Computation Wiki" (2018) https://news.ycombinator.com/item?id=18146854 :

> "Quantum knowledge cools computers: New understanding of entropy" (2011) https://www.sciencedaily.com/releases/2011/06/110601134300.h...

>> The new study revisits Landauer's principle for cases when the values of the bits to be deleted may be known. (with QC)

https://en.wikipedia.org/wiki/Landauer%27s_principle

The heat can be reduced by factor of a billion or so.

Reversible circuits [1] built with superconductors could generate no excess heat at all.

[1]: https://spectrum.ieee.org/the-future-of-computing-depends-on...

We won’t be getting long distance high voltage electricity transmission lines made of this (possible) room temperature superconductor for the same reasons we don’t have high voltage long distance electricity transmission made of gold, a much better conductor than copper.

What things are we going to do with this material in the next 5 years? What things that I buy and use will be made better?

You know those joke buzzer/shocker rings? Well, they're going to get really, really good.

Just what I've always wanted!

Hoverboards. I want those damn things as was prophesied by back to the future.

>The ramifications of the inflection point we are currently at is mind boggling.

No joshing, I just got done posting this on facebook: "If this is practical it will change the world like the transistor; maybe more."

Ha, I just posted about a dragonfly I saw while walking, and how it seemed unperturbed by the possibility of room temperature superconductors on Earth. I added that I should probably learn from the dragonfly.

Have you ever seen a dragonfly sit still? I'm just saying, I wouldn't describe them as unperturbed.

I'd be terrified if I was the dragonfly. Its life and the lives of all other living beings depend more on what we do with our technology than on anything else.

Sounds more like the perennial "10 years in the future" techs, like memsistors, cold fusion, holographic storage, and so on.

Reply on news.ycombinator.com