On Sun, Jan 13, 2013 at 10:39 AM, Dawid Weiss <[email protected]> wrote: > I doubt it'll be faster,
Yeah, this shift will be pretty simple - 1 cycle, and normally with a max throughput of 2 to 3 per cycle for x86 server CPUs. Even without java bounds checking, the most you could hope for would be to match the performance with a lookup. Also, L1 access is still greater than 1 cycle on a number of processors. -Yonik http://lucidworks.com --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
