Re: Overclocking an Apple ][?

[mjmahon@aol.com (Michael J. Mahon)]

Simon wrote:

>Michael J. Mahon <mjmahon@aol.com> wrote:
>
>> The empty FOR loop is a better test, since it runs a lot of
>> code from ROM and only updates a few interpreter variables.
>> 
>> In general, the part of a program that changes the screen will
>> run unaccelerated, while the part that simply computes with
>> relatively few stores will experience maximum acceleration.
>
>Which means, in practice, that some programs will actually run slower on
>the IIc+ @ 4MHz than on a IIgs :(

Yes, if the IIgs writes to memory at the higher speed, which I
think it does, and then "shadows" the video data to slow RAM.

>> If you were taking the square root of I and then printing,
>> you would notice the difference even with the scrolling
>> overhead.
>
>Has there ever been any diagnostic software that can determine system
>speed, or does the Apple II's hardware preclude such a thing?

What's historically missing is a real-time reference.  But with
the VBL sensing of the //e and later machines, a speed test
could easily be written.

All that would be required is to wait for VBL, then start counting
in a tight M/L loop (using registers) while waiting for the next VBL.
The count is the number of iterations of the loop in 1/60th of a
second.

By inserting memory operations you could also determine the
memory write speed, and by indexing over a range of memory,
you could determine the cache size!

-michael

Check out amazing quality sound for 8-bit Apples on my
Home page:  http://members.aol.com/MJMahon/