By default, all computations occur in hardward double precision. While this precision could be large enough to obtain accurate results, for larger problems, one may need to increase the precision to double double or to quad double precision.

Setting the value of computingPrecision to 2 changes the precision in the path trackers to double double.

Setting the value of computingPrecision to 4 changes the precision in the path trackers to quad double.

To compensate for the cost overhead of the higher precision, it is useful to run the multithreaded versions of the path trackers, see the option numThreads.

- solveSystem(..., computingPrecision => ...) -- option to specify the working precision

The object computingPrecision is a symbol.