Hades, memory access etc

From: Petr Stehlik <stehlik_at_cas3.zlin.vutbr.cz>
Date: Tue, 13 Jan 1998 15:16:42 +0100

Hi Magnus and all,

I remember saying that my friend's Hades was much slower than my
Afterburner. And yes, TOSQuake was much slower on his machine than on mine

However in the meantime I got another GEMbench and Nembench figures, which
proved that 'normal' Hades60 is much faster in memory access than
Afterburner. That second Hades was running at 68.5 MHz, IIRC. Anyway, the
reason for too low GEMbench and Nembench numbers of tests on the first Hades
was simple - the caches were surely turned OFF.

The 68060 in Hades is driven by strange CPUxxx.ACC with a text file
containing list of programs and cache flags for every listed program. I bet
that by default the caches are off (or something similar) so the tests were
running with caches turned off.

If this theory is not right, I have no other idea why one Hades60 is about
ten times slower than others.

BTW, that second Hades compiles MiNT in two minutes, while my AB needs about
7 (iirc).


P.S. Anyone tried my Atari800 emulator? Yesterday I added there a speed
indicator - which indicates, that the emulator is about four times slower
than original machine running at 1.79MHz! :-(

E-mail: stehlik_at_cas3.zlin.vutbr.cz                        PARCP developer
   WWW: http://ft3.zlin.vutbr.cz/stehlik/home.htm         MiNTOS/Linux user
mirror: http://www.cyberstrider.org/mirrors/stehlik/      Atari Falcon040
Received on ti. jan. 13 1998 - 20:41:00 CET

This archive was generated by hypermail 2.3.0 : ti. nov. 03 2015 - 20:07:53 CET