Re: How to benchmark a modified ZipGSX?

["Sean Fahey" <a2fan@hotmail.com>]

"Dan Foster" <usenet@evilphb.org> wrote in message
slrndgs5om.4gu.usenet@zappy.catbert.org">news:slrndgs5om.4gu.usenet@zappy.catbert.org...

> I'm trying to determine what speed the GS is actually running at, with a
> modified ZipGSX installed.

If I understand you correctly, a test of that nature might be a little
misleading. As you're probably aware, the speedup derived from a ZipGSX
isn't just from it's raw clock speed, but how much cache is installed and
how efficiently it's working as well. And how many programs are running, how
much I/O is going on, etc. Lots of variables that a straight test from
ProDOS wouldn't be able to take into account.

I'm not aware of an NDA out there that can tell you your ZipGSX cache hit
efficiency.

If you're just trying to find out how fast your Zip is, take your oscillator
speed and divide by 4 - so a 32MHz oscillator is a 8MHz ZipGSX.