I have been doing some thinking and this is game changing but not so much. We won’t get hoverboards or flying cars to my knowledge. We will get much cheaper maglev trains, but in America we refuse to build public infrastructure that isn’t for cars so that isn’t gonna fix it.
We won’t get faster traditional computers because those need semi-conductors. There are some patents and theories about superconducting transistors so we may get a “cool running” cpu eventually, but it won’t be faster it just won’t heat up.
Quantum computers will get cheaper and maybe more available, but they are still a research topic so we are probably decades away from them having practical use (or ever in terms of practical for everyday use, they will break encryption as we know it though).
We will “instantly” save something like 30% of our power generation that is lost to heat, but again that is going to require a massive infrastructure project to replace all high voltage power lines, so that is never going to happen in America.
Brush less motors will be able to be smaller and/or able to take in more energy so they will be more efficient, but we are still beholden to our energy storage density.
There is a theoretical idea of using superconducting rings to let electrons flow around it indefinitely, as an energy storage medium, but I have no idea how close that could be or how dense that would be compared to Lithium Ion batteries, or Fossil fuels which is the real competition.
We will get smaller and cheaper MRI’s so medical imaging should get cheaper and more available to the “global south”.
More heat efficient processors and more energy efficient processors are one and the same. Which is huge.
Energy usage is a large portion of the cost of computational infrastructure, and things like training neural networks.
I suspect a thermally more efficient processor would also potentially last much longer too, with less intense thermal cycling.
A lot of data centers are limited by the energy infrastructure where they are constructed.
I thought heat is the main thing limiting computer performance? Like, if we had superconducting transistors that take little energy to change state, highly parallel tasks that are power-limited today would get a whole lot faster. Think native 4k path tracing-level graphics in games on our phones. And better/faster/cheaper AI systems, though they are limited more by memory than by compute, so they’d likely still be run in the cloud mostly.
Heat is a big issue, but we are close to the physical limits of transistor size, they are nearly the size of atoms AFAIK.
So this will allow us to have more of them closer I guess with no heat limits. There is also a lot of stuff that goes above my head about quantum tunneling when our transistors get that size. But transistors use semiconductors (Sillicon) not conductors, so this isn’t a drop in replacement. Will require a new type of transistor that uses a conductor I suppose.
I have been doing some thinking and this is game changing but not so much. We won’t get hoverboards or flying cars to my knowledge. We will get much cheaper maglev trains, but in America we refuse to build public infrastructure that isn’t for cars so that isn’t gonna fix it.
We won’t get faster traditional computers because those need semi-conductors. There are some patents and theories about superconducting transistors so we may get a “cool running” cpu eventually, but it won’t be faster it just won’t heat up.
Quantum computers will get cheaper and maybe more available, but they are still a research topic so we are probably decades away from them having practical use (or ever in terms of practical for everyday use, they will break encryption as we know it though).
We will “instantly” save something like 30% of our power generation that is lost to heat, but again that is going to require a massive infrastructure project to replace all high voltage power lines, so that is never going to happen in America.
Brush less motors will be able to be smaller and/or able to take in more energy so they will be more efficient, but we are still beholden to our energy storage density.
There is a theoretical idea of using superconducting rings to let electrons flow around it indefinitely, as an energy storage medium, but I have no idea how close that could be or how dense that would be compared to Lithium Ion batteries, or Fossil fuels which is the real competition.
We will get smaller and cheaper MRI’s so medical imaging should get cheaper and more available to the “global south”.
Am I missing anything?
More heat efficient processors and more energy efficient processors are one and the same. Which is huge. Energy usage is a large portion of the cost of computational infrastructure, and things like training neural networks. I suspect a thermally more efficient processor would also potentially last much longer too, with less intense thermal cycling.
A lot of data centers are limited by the energy infrastructure where they are constructed.
I thought heat is the main thing limiting computer performance? Like, if we had superconducting transistors that take little energy to change state, highly parallel tasks that are power-limited today would get a whole lot faster. Think native 4k path tracing-level graphics in games on our phones. And better/faster/cheaper AI systems, though they are limited more by memory than by compute, so they’d likely still be run in the cloud mostly.
Heat is a big issue, but we are close to the physical limits of transistor size, they are nearly the size of atoms AFAIK. So this will allow us to have more of them closer I guess with no heat limits. There is also a lot of stuff that goes above my head about quantum tunneling when our transistors get that size. But transistors use semiconductors (Sillicon) not conductors, so this isn’t a drop in replacement. Will require a new type of transistor that uses a conductor I suppose.