Untested, but it should work:

use strict;
use warnings;

my ($in, $out, %f1);

open($in, 'in1.dat');
while (<$in>) {
    chomp; $f1{$_} = ();
}
close($in);

open($out, '>out.dat');
open($in, 'in2.dat');
while (<$in>) {
    print $out if !exists $f1{(split /\t/)[0]};
}
close($in);
close($out);
[download]

I assumed your delimiter was \t. Go ahead and change that if I'm wrong.

I'm also assuming you don't have to run this a bunch of times simultaneously. If memory usage is a problem, you can just step through the files one line at a time and compare (since the files appear to be sorted already). But this should be easily doable with a hash, even in Perl.

EDIT: By "this" I meant the task at hand, not walking through the files one line at a time and comparing. Just thought I'd clarify my bad writing :)