Is it possible to make a crypto mining code with python that is solely based on DIMM speeds?

[Skept]

[Helpful Tech Witch]

 

 

 

 

oh wait this isnt post random emojis?

Well then, maybe, but I dont think you understand how crypto works. Plus, I feel like this could wear out ram, because you would proably be writing to it and unwriting in the next clock to prove the ram is runnign at the speed it should be

[Skept]

1 minute ago, HelpfulTechWizard said:

Alot of people have excess ram so ppl will definitely mine it.

 

 

 

 

oh wait this isnt post random emojis?

Well then, maybe, but I dont think you understand how crypto works. Plus, I feel like this could wear out ram, because you would proably be writing to it and unwriting in the next clock to prove the ram is runnign at the speed it should be

 

[Eigenvektor]

Cryptomining has to "produce" something, some kind of unique token (or "coin"). Mining a coin must take a certain amount of computational effort otherwise anyone can just generate however many tokens they want, making it worthless.

 

As a simplified example, Bitcoin computes SHA-256 hashes. A hash is only a valid coin, when that hash meets certain criteria (e.g. starts with x zeroes). Since a hash is irreversible, you can't just generate a coin and then perform the operation in reverse. This means you have to test millions or billions of numbers to find one that produces a winning hash. When computational power increases, the difficulty is increased by making the criteria more strict.

 

If you can come up with an algorithm that somehow benefits from lots of RAM or fast RAM, then sure, you can use that for some form of cryptocurrency. The programing language is secondary, you need to come up with an algorithm first (and I'd assume for the sake of efficiency, most people would not pick Python to implement it).

[porina]

Don't know about any limitations that Python might have, but in a very generic sense I have to wonder if this might be achieved by something similar to ethash but with two changes: 1, use a data set, thus the amount of ram required, that exceeds that of common GPUs to prevent their use, and 2, use a computation method that can't easily be implemented by GPU but can be by CPU.

 

Part 1: might be difficult now, given that 16GB is slowly moving mainstream in GPUs. 32GB systems would be the next logical step, less a bit for OS and other uses. Some high end GPUs might still achieve this. Excluding 16GB ram systems may reduce interest.

 

Part 2: might override part 1. GPUs are still best at stuff that scales wide, and you'd have to find a way to come up with a hashing workload that doesn't scale on GPU, even by running multiple instances. If the goal is to use ram perf as the limiting factor, then it would have to be capable of using that ram perf.

[Skept]

17 minutes ago, Eigenvektor said:

Cryptomining has to "produce" something, some kind of unique token (or "coin"). Mining a coin must take a certain amount of computational effort otherwise anyone can just generate however many tokens they want, making it worthless.

 

As a simplified example, Bitcoin computes SHA-256 hashes. A hash is only a valid coin, when that hash meets certain criteria (e.g. starts with x zeroes). Since a hash is irreversible, you can't just generate a coin and then perform the operation in reverse. This means you have to test millions or billions of numbers to find one that produces a winning hash. When computational power increases, the difficulty is increased by making the criteria more strict.

 

If you can come up with an algorithm that somehow benefits from lots of RAM or fast RAM, then sure, you can use that for some form of cryptocurrency. The programing language is secondary, you need to come up with an algorithm first (and I'd assume for the sake of efficiency, most people would not pick Python to implement it).

what programming software do u think would be applicable (I am really thinking about pursuing this idea)

[Eigenvektor]

Just now, Skept said:

what programming software do u think would be applicable (I am really thinking about pursuing this idea)

As I said, programming language is a secondary concern. Worry about coming up with an algorithm first, then pick whatever programming language is most efficient for implementing it.

[Skept]

Just now, Eigenvektor said:

As I said, programming language is a secondary concern. Worry about coming up with an algorithm first, then pick whatever programming language is most efficient for implementing it.

I'm a complete newby so I guess I'll just start learning about algorithm's 

 

[tikker]

Also realise that while the time of random (shit)coins pumping is most certainly not over, if you want your coin to be worth something at all or see it being used besides just hype (think shiba, doge etc.) then you need a use case. What is your coin going to do that the hundreds or thousands of others don't already do? How will it be better and/or more efficient at it? Why would people use your coin over existing ones? Will your solution scale properly should it see large(r) scale usage? All things to consider.