clearwater304
- 88
- 0
For instance, the largest known prime number has nearly 13 million decimal digits. Would any normal computer be able to calculate this?
Awsome. I know when I took java a few years ago a lot of software would only store numbers up to a certain number of digits
clearwater304 said:Is there a way I can run mathematica remotely on their server.
I'm don't think that you meant this, but it should be reiterated that one does not ever need to find a factor of a number to prove that it's composite, even for extraordinarily large numbers.phyzguy said:What are you trying to do? Calculating a number like this is much, much faster than proving that it is prime. Encryption algorithms rely on the fact that multiplying two large primes to generate a large composite is much (much) faster than extracting these two primes from the resulting composite.
Solving 10^13000000=2^x gives x~40000000, which means it will take about 40 million bits to represent the number, which would be 5 million bytes. Since most computers have on the order of billions of bytes (gigabytes) the memory isn't the problem. The problem is the computation.clearwater304 said:Is there a way I can run mathematica remotely on their server. If I try to do a primality test on a very large number on my notebook it gives me a recursive error. I suppose this is due to the fact that it ran out of memory and I only have 4gb on my notebook.
Good luck with that. You might want to look at some other the existing ones first. The AKS primality test is the best known primality test.clearwater304 said:Mathematica seems to be the way to go, now I just have to create a customized primality test so it doesn't take a year to prove.
You do it you want your algorithm to be anywhere near efficient. EDIT: Never mind, I was wrong. I was presuming you were referring to the Wilson test, but I just remembered the quadratic residue test, which is a very efficient way to eliminate a lot of composite numbers without finding a factor.Feryll said:I'm don't think that you meant this, but it should be reiterated that one does not ever need to find a factor of a number to prove that it's composite, even for extraordinarily large numbers.
GMP, a C and C++ library, can handle arbitrarily large numbers and is very heavily optimized. So can Java's BigInteger, but I don't know how optimized it is.clearwater304 said:For instance, the largest known prime number has nearly 13 million decimal digits. Would any normal computer be able to calculate this?