c# - Parallelizing very large array base conversion -

i have method converts value newbase number of length length.

the logic in english is:

if calculated every possible combination of numbers 0 (c-1) length of x set occur @ point 

while method below work perfectly, because large numbers used, can take long time complete:

for example, value=(((65536^480000)-1)/2), newbase=(65536), length=(480000) takes hour complete on 64 bit architecture, quad core pc).

private int[] getvalues(biginteger value, int newbase, int length) {     stack<int> result = new stack<int>();      while (value > 0)     {         result.push((int)(value % newbase));          if (value < newbase)             value = 0;         else             value = value / newbase;     }      (var = result.count; < length; i++)     {         result.push(0);     }      return result.toarray(); } 

my question is, how can change method allow multiple threads work out part of number?

i working c#, if you're not familiar pseudocode fine too.

note: method question: cartesian product subset returning set of 0

if getvalues method bottleneck, there several things can speed up.

first, you're dividing newbase every time through loop. since newbase int, , biginteger divide method divides biginteger, you're potentially incurring cost of int-to-biginteger conversion on every iteration. might consider:

biginteger bignewbase = newbase; 

also, can cut number of divides in half calling divrem:

while (value > 0) {     biginteger rem;     value = biginteger.divrem(value, bignewbase, out rem);     result.push((int)rem); } 

one other optimization, mentioned in comments, store digits in pre-allocated array. you'll have call array.reverse them in proper order, takes approximately no time.

that method, way, doesn't lend parallelizing because computing each digit depends on computation of previous digit.