For example:

#!/usr/local/bin/perl -w
use bignum;

my $num1 = 4;
my $num2 = 5;
my $mult = $num1 * $num2;
 
print sqrt($mult),"\n"; # Yup

For a number that would be large, you can use bignum's accuracy option, a, to set the number of places after the decimal:

perl -Mbignum=a,50 -le 'print sqrt(20), "\n";'

If you wish to use rounding with bignum, you can use the precision option, p, with a 0 or 1 value to round up a value:

perl -Mbignum=p,0 -le 'print sqrt(24), "\n";' # Gives 5

To enable tracing for bignum, use the t or trace options:

/usr/local/perl5.8-prerc1/bin/perl5.7.3 -Mbignum=p,0,t -le 'print sqrt(24)'
MBI import Math::BigInt::Trace :constant upgrade Math::BigFloat::Trace lib Calc
MBF import Math::BigFloat::Trace :constant downgrade Math::BigInt::Trace
MBI new '24' => '24' (Math::BigInt::Trace)
MBF new '24' => '24' (Math::BigFloat)
MBF new '4' => '4' (Math::BigFloat)
MBF new '1E-4' => '0' (Math::BigFloat)
MBI new '2' => '2' (Math::BigInt::Trace)
MBF new '2' => '2' (Math::BigInt::Trace)
MBI new '38' => '38' (Math::BigInt::Trace)
MBI new '38' => '38' (Math::BigInt::Trace)

The l, or lib, option lets you specify a different math library:

perl -Mbignum=l,new_math_lib -e 'print $this ** $that'

The v, or version, option shows you the version of all modules used by bignum and then exits.

8.26. bigint		8.28. bigrat