4 ms·
Ask HN: If LK-99 works, will existing crypto security be at risk?
Assuming we manage to use the superconductor to create computers 1000x more efficient, or even quantum computers. Will this "break" existing crypto, for example, due to the suddenly insanely high hash rates? Or even by "encryption-breaking" mechanisms?
- mikequinlan 3y agoHow would room temperature superconductors magically make silicon semiconductor CPUs run faster?
- floxy 3y agoYou'd make Josephson Junction based microprocessors with superconductors (replacing the FET based CMOS architectures, and get rid of the silicon completely), and that might lead to a ~100x speed increase. And then the low power could enable integration in the third dimension, where you can stack active circuitry on top of each other, leading to another 10x in density. Right now semiconductors transistors are limited to living on the silicon substrate. There are efforts at doing silicon-on-insulator transistors, but heat dissipation has been a limiting factor.
- mikequinlan 3y agoBut how would room temperature superconductors over a lead-based material like LK-99 lead to any of this?
- floxy 3y agohttps://en.wikipedia.org/wiki/Superconducting_computing https://en.wikipedia.org/wiki/Superconducting_computing
- mikequinlan 3y agoThat article talks about cryogenic superconducting as is used in current quantum computers. Allowing room temperatures would greatly reduce costs but I don't see how it speeds anything up. But even then, the current discovery only works with the LK-99 alloy and not any of the materials used in superconducting logic.
- floxy 3y agoYes, the textbooks will need to be rewritten to remove references to cryogenic temperatures if we have a breakthrough technology in room temperature superconductors.
- jacquesm 3y agoBecause it cuts the gate charge time to essentially zero.
- tromp 3y agoAbility to compute hashes orders of magnitude faster will no more break crypto than the evolution from CPU mining to GPU mining to FPGA mining to ASIC mining did. So, no that part doesn't break crypto. On the cryptocurrency side, the difficulty self adjusts. On the the password cracking side, it reaffirms the need to use memory-hard hash functions that don't speed up so easily.