In the past year or so we have seen production of ASIC devices designed for mining of cryptocurrencies. These devices can perform SHA256 hashing at rates much higher than was seen in the past and are continually advancing in power. Can such devices be used to perform cryptanalysis instead of mining? Does this development pose a threat to our current security assumptions?
Asked
Active
Viewed 1,961 times
1 Answers
2
Practicality: Using a bitcoin miner for cryptanalysis would at the very least require you to write very low-level custom code. Indeed, depending on the precise hardware/software split used by the miner, it might well require modifying the actual hardware to facilitate your cryptanalytic attack. Anyway, let's suppose someone could 'convert' one. What would this mean? Well, if someone could convert a bitcoin ASIC into a viable cryptanalysis box, then you can bet the ASIC companies would just go ahead and produce them directly.
It is probably reasonable to assume that 'groups actively involved in cryptanalytic attacks already use some of the best hardware available, so in practice a converted miner would still probably leave you behind the curve. Anyway, supposing it did, your question would boil down to:
What would happen if the 'effectiveness'$^{[1]}$ of cryptanalytic equipment suddenly advanced by an order of magnitude?
Most good crypto is either believed to be well outside the bounds of brute force (eg see these answers about AES, RSA), even when we make very generous assumptions about an adversaries computational power. As such, even relatively significant improvement in cryptanalytic equipment (which is what your question suggests) will make a negligible difference in these problems.
Notes:
- We assume 'effectiveness' has some appropriate meaning for the problem in hand, eg cost/size/speed etc
Cryptographeur
- 4,357
- 2
- 29
- 40