To determine the error-correcting capability of a large error-correcting code it may be necessary to generate the code, an intractable task. Using a stack-based algorithm and utilizing structural properties of a code can reduce the time required. Timing results are reported for generating large codes using these methods on massively parallel platforms.