Re: ramfast-ramkeeper-transwarp combo question

[a2aviator@gmail.com]

Here's my take;

Some comments in a nutshell variety.

But first- a little background history.  The currently available CPU
from WDC, I usually refer to as the "Sanyo CPU".  Sanyo, as a second-
source licensee - manufactured the 65816 to a finer spec and made
changes to support a better yield as well as take care of some issues
within the design.

When the accelerators were being designed, this being the MD Ideas
product, the Transwarp GS and the ZipGSX.

Each company had their own way for doing it. The ZipGS and MDIdeas are
similar concepts at the loosest point. Each design had to work around
design flaws and anomalies. Thats the reason for all the GAL revisions
on the TWGS.

Here's how I categorize each of the two:

The Zip GS is a fast CPU replacement- a hardware solution. As you
know, the IIgs fails a few of the inbuilt test routines.

The TWGS is a "compatible" solution that goes out of it's way to make
sure the user sees no difference except speed.

The Zip GS has dip switches where the user can specify the experience
regardless of outcome.

The TWGS will for the most part, not allow anything that could produce
unpredictable or undesirable results to happen.

Both cards do equally as well in the end.

In the beginning, I had a TWGS for obvious reasons, thats all you
could get. In fact, I had one a few months before they were released
for compatibility checking with the company I worked for at the time.

Similarly, with the Zip GS, I had access to it a bit early.

My first TWGS was 8MHz and it was .. well, it worked for the most
part. There were some issues. AE would send GALs and CPUs at times to
swap stuff out. At one point the card was going to go back and I
convinced 'em not to because I have this thing for having early
production items. :) What AE was doing with the GAL revisions was
making adjustments for reactions and speed regulation.

The 65C816 had some issues with certain instructions, timing and what
not that meant sometimes when things happened in certain sequences or
before/after certain states - the speed had to be slow, or at least
slowed down.

The Sanyo manufactured CPU fixed all of this.

This is why I've always used the Rev. A GAL set with the TWGS in
dealing with speed upgrades. Since this is considered to be the least
restrictive.

I have flip-flopped on them over the years to where now it's pretty
much either or, but from a marketing viewpoint. If you have lots of AE
hardware and are concerned, you can be sure that the AE products all
work with the other AE products. The TWGS will provide a more
transparent environment, the Zip GSX will provide what it says but you
may have some slight issues in dealing with niche products and
situations.

Though the latter is also possible with the TWGS.

Which one is better, right, etc?

The one you can get your hands on.  Given a choice with two in front
of me, if I could leave the table with one - if I could not program
GALs, and the TWGS did not have Rev. A GALs on it... I'd take the Zip
GSX.

If it did, I'd still be there trying to decide. It's a toss-up.

The Zip GSX is faster, the TransWarp GS is more cautious. When
comparing the same clock speed.

--
tdiaz(a-t)apple2(dotsero)0rg