(untitled)

list · 2021-10-10 17:07

Searching a BRUTE FORCE problem, or something similar. I have some CPU cycles to burn on a Raspberry Pi cluster. Less than 100 devices, around 6 months to use. The cluster is not connected to Internet. Any ideas?

Elliot Alderson · 2021-10-10 17:52

Seriously? You cannot think of *any* bruteforce problems?

Just another Random Lainon · 2021-10-10 19:50

https://en.wikipedia.org/wiki/Infinite_monkey_theorem

pop · 2021-10-11 02:28

>>27884 There are several problems in this category from password cracking to 3N+1 search. Off-line nature limits the operation somewhat. >>27886 Although there are more nodes than typical small-scale proof-of-concept parallel computation platforms, the actual computing power is quite limited for infinite problems. There would be certain poetic beauty in Random Shakespeare, but this will probably not get attempted. So, any other ideas for inspiration?

Lunch · 2021-10-11 05:34

Look a bit into Matrix Profiles, they're a crazy efficient and easy to parallelize time series analyzer. All you need is a big pile of time series data and you can do some pretty deep analysis.

push · 2021-10-11 05:35

>>27896 Here, for your consideration... https://youtu.be/1ZHW977t070

dictionary · 2021-10-11 06:47

>>27896 >big pile of time series data Massive datasets are rather out of the question. Platform consists of Raspberry Pi hardware and there is no planned data repository or massive server capacity. Large data? I might provide capacity, but I cannot immediately identify any interesting time series or other 1-dimensional material of interest. Interesting concept, though. Matching image patterns is on my list, but my targets are 2-dimensional features below noise level. Nevertheless, this warrants otherwise enough interest that I will click the second part of the video as well, so thanks!

Mario Rossi · 2021-12-12 00:53

Are there programs that can work out the optimal windings for an electric motor without me having to do much of the work myself? I want to know how lightweight or hand-windable a design I can use to run reliably inside a 250 GTO replica chassis

Turing · 2021-12-12 01:20

Here's something: http://www.ii.uib.no/~osvik/pub/aes3.pdf Calculate for me a more efficient method of calculating Serpent Cipher S boxes and their inverse boxes. I don't see a reason why any of the sixteen boolean functions shouldn't be used in the search. I'd appreciate it.