0
I see this kind of question before on this topics, but lots of them seems very old. Arcording to Moore's law, computer speed doubles every two year, so i wonder how fast have they archive these days. Can supercomputer capable of doing such a thing?
1
My laptop, running an i7-4XXX can hit about 1 millions keys per second using a vanity address generator at its peak. This usually slows down as it heats up, and it is certainly possible to go faster with GPUs.
Assuming the strongest hardware today is 10000x faster, or can generate 10 billion keys per second, you will reach 10^18 in 10^18/10^10 seconds, which is 1157 days.
Note that the Bitcoin keyspace is much larger, about 2^256 (although I believe you "only" need to crack 2^128 to start breaking into stuff with high probability, but I can't find that reference right now). For 2^256 keys, even at 10 billion per second, it would take you 3.672×10^59 years, which is 2.7×10^49 times the age of the universe.
Edit:
You would only need to go through 2^160 combinations to break a p2pkh address, since you only need to beat the HASH160. This reduces your time to only 4.631×10^30 years.
my GTX 780 Ti run on it can make nearly 60mil keys per seconds. btw no need to hash the entire 2^256 combinations, i just want to find out if any hardware out there can hash the number i say in fastest time possible – Huang Lee – 2018-08-07T16:08:29.407
Generating an address involves more than just generating a key. You need to generate the private key, public key, and then hash160 it, and then base58 encode it. All of these operations slow down the process. – Raghav Sood – 2018-08-07T16:10:01.320
According to this thread, a 1080ti can manage ~100 million addresses per second
what about supercomputer? i see lots of ppl suggest this but didn't find anybody try any experiments on it – Huang Lee – 2018-08-07T16:29:06.580
I don't know of any numbers for those platforms, and with the numbers involved in bruteforcing keys, it's simply not worth it for someone to even experiment on multimillion dollar machines, because they still will not get anywhere – Raghav Sood – 2018-08-07T16:39:24.250
Why 10^18? I can't see any logic or significance to that number – Raghav Sood – 2018-08-07T15:54:37.907
@RaghavSood it about amount of 64bits possible passwords combinations, lowest form of security for cryptocurrency i found, so this number might suitable to try – Huang Lee – 2018-08-07T15:58:01.810
2^64 is still about 18.5x larger than 10^18 – Raghav Sood – 2018-08-07T16:13:02.757
"Strongest hardware" is kind of impossible to define since this task is totally parallelizable. Whatever hardware you have in mind, just get two of them, and then you'll finish the job in half the time. – Nate Eldredge – 2018-08-07T16:25:18.380