exxos's Atari, Amiga & retro forum

As said above the CPU can't access ST-RAM faster than the blitter anyway. It may show a small speed up, but that would be the 030 caches getting involved.

Its probably not so simple with TOS routines or even any blitter "effects" involved. I mean a inversion on the blitter or any sort of processing could take a few bus cycles worth of time, the CPU could do those things quicker. I don't know how easy that would be to test overall. Though I guess the CPU would probably win overall, especially with caches turned on anyway.

EDIT:

In case of GB6, its only doing RAM to RAM copying.. no effects.. probably why it shows 190% speed.. on that result, its twice as fast as the CPU.. but its not so simple because TOS's routines are going to be way slower running on the CPU anyway.

EDIT2:

Also bare in mind, should we get the suska core running as we want it, we will have a faster blitter running at 50mhz or whatever, so don't dismiss the blitter just yet ;)

In terms of fast-ram, the suska core will simply have 14MB ST-RAM running at 50MHz. So fast-ram basically becomes obsolete then. Thus no need to even think about the blitter having access to fast-ram.. as well.. it should have access to the full 16MB map or whatever anyway.. if it does not, then we will make it do it ;)

If it involves any questions about the Pak specific TOS 3.06 modified version, you
guys should really get Holger Zimmermann involved in this conversation. The man
is amazing.

When I first got my Pak board with no TOS version and I contacted him, he was
happy to make an English version that would work on my STacy. Sent it to me
by e-mail and refused any compensation (I tried).

He goes by "pakman" on AtariAge and AtariForum, IIRC. I think I also have his
e-mail address lying around somewhere, although I haven't talked to him directly
in quite some time.

PS I think I have the pak TOS notes somewhere as well.

exxos wrote: 08 Oct 2020 15:56
I know its been asked before.. Though IIRC I verified @Darklord PAK 030 TOS with proper TOS and it was the same.. Though I know I did the same with my PAK but it was a 020 version which came from @frank.lukas . I think Darklord thought it was patched but it actually wasn't.. but can't remember now exactly as was years ago.. maybe someone with pak030 306 can just verify if its the same as generic 306 or not...

I honestly don't remember. I don't know how they could be *exactly* the same though. It's just visual but the info screen and other things
are changed. Here is an excerpt from an e-mail that I exchanged with Holger, right before he sent me the modified version. Seems there
is something to do with the boot loading process, but I'm not sure about the TT video and blitter routines that have been mentioned.

As I posted elsewhere, really need to get Holger involved. :)

Hi Ronald,

Some hints:
All these combinations work fine:
TOS 2.06 on the Atari mainboard, no ROMs on the PAK, or ROMs on the PAK disabled
TOS 2.06 on the Atari mainboard, TOS 2.06 (4 ROMs, no patches required) on the PAK
TOS 2.06 on the Atari mainboard, TOS 3.06 (4 ROMs, with PAK patches) on the PAK

With TOS 1.xx on the Atari mainboard things are different:
After reset the ROM on the mainboard tell the CPU to jump to $FC0030, which is ok for TOS 1.xx.

But with TOS 2.06 or 3.06 we need a jump to $E00030.

In the PAK-patches for TOS 3.06 a so called boot-patch is included, which does the job.

But this patch is only tested for the German version so far.

Anyhow, I am going to compile an english version, with this patch included, and then let us see.

US or UK version?

Regards, Holger

[UNQUOTE]

Double post, apologies...