Page 11 of 110
Re: REV 3 - The beginning
Posted: 01 Apr 2021 18:05
by exxos
More results, no idea if these are good or not..
IMG_6359.JPG
Code: Select all
NemBench v2.1 - precision CPU/FPU profiler.
Integer multiply (16bit) -> 1.921 Mips (~313%)
Integer divide (16bit) -> 1.135 Mips (~313%)
Linear (stalled) integer -> 24.899 Mips (~312%)
Interleaved (piped) integer -> 24.975 Mips (~313%)
16bit read (100% hit) -> 24.630 MByte/sec (~313%)
16bit write (100% hit) -> 16.835 MByte/sec (~279%)
32bit read (100% hit) -> 49.261 MByte/sec (~313%)
32bit write (100% hit) -> 33.557 MByte/sec (~503%)
Linear 32bit read (ST-Ram) -> 3.919 MByte/sec (~73%)
Linear 32bit write (ST-Ram) -> 3.919 MByte/sec (~60%)
Linear 32bit copy (ST-Ram) -> 1.964 MByte/sec (~60%)
Linear 32bit read (FastRAM) -> 32.564 MByte/sec (~612%)
Linear 32bit write (FastRAM) -> 32.403 MByte/sec (~502%)
Linear 32bit copy (FastRAM) -> 14.419 MByte/sec (~446%)
Re: REV 3 - The beginning
Posted: 01 Apr 2021 19:25
by Badwolf
exxos wrote: 01 Apr 2021 18:05
More results, no idea if these are good or not..
In a global perspective? Very (better than my Falcon booster, FWIW).
Versus other 536s? Dunno. But your Gembench AltRAM and Integer Div scores are now spot on.
:thumbup:
BW
Re: REV 3 - The beginning
Posted: 01 Apr 2021 20:18
by exxos
Badwolf wrote: 01 Apr 2021 19:25
In a global perspective? Very (better than my Falcon booster, FWIW).
Need to get you running now ;)
Amazingly space Harrier loads on the music works... On my SEC booster I had to have a separate decoding line for the audio, and yet I ported a lot of to the 536 anyway :shrug:
Re: REV 3 - The beginning
Posted: 01 Apr 2021 20:23
by exxos
Got off level 1!! , did not get off level 2 :lol: :roll:
IMG_6361.JPG
Xenon didn't fare too well :lol:
IMG_6362.JPG
Castlemaster still runs like shite :lol: :roll:
Re: REV 3 - The beginning
Posted: 01 Apr 2021 21:10
by Badwolf
exxos wrote: 01 Apr 2021 20:18
Need to get you running now ;)
I couldn't make the 100MHz clock work so my RAM's clocked the same as the CPU. Plus I only have two lines per burst.
That said, it is *almost* as quick.
Building my second Rev3 board in the hope of letting it into the wild. Chasing a bloody rogue data line. Single access: fine. YAARTTT? D11 always high. Problem seems to be on the 3V3 side so not pull up related, as far as I can see. Every other line fine. What's that all about, eh?
Amazingly space Harrier loads on the music works... On my SEC booster I had to have a separate decoding line for the audio, and yet I ported a lot of to the 536 anyway :shrug:
:cheers:
We need a list of games that a) run with an 030 and b) use AltRAM.
1) Frontier. Obvs.
2) Err...
There's a lot of SDL stuff that should work -- PM-Doom's a staple for me -- but it (and I suspect most) assume 256 colours. :-(
BW
Re: REV 3 - The beginning
Posted: 01 Apr 2021 21:15
by exxos
Badwolf wrote: 01 Apr 2021 21:10
Building my second Rev3 board in the hope of letting it into the wild. Chasing a bloody rogue data line. Single access: fine. YAARTTT? D11 always high. Problem seems to be on the 3V3 side so not pull up related, as far as I can see. Every other line fine. What's that all about, eh?
Not noise is it ? I had a bonkers issue with D4 not long ago. In only worked with the scope probe on it :lol: :roll:
Badwolf wrote: 01 Apr 2021 21:10
We need a list of games that a) run with an 030 and b) use AltRAM.
Yeah, people have been saying the same for years, but nobody ever compiles a list.
Re: REV 3 - The beginning
Posted: 01 Apr 2021 22:08
by Badwolf
exxos wrote: 01 Apr 2021 21:15
Not noise is it ? I had a bonkers issue with D4 not long ago. In only worked with the scope probe on it :lol: :roll:
Wouldn't rule it out yet but all the other lines, including ones adjacent, seem OK. I've had the buffer chip off and symptoms are the same so I'm going to work on the RAM first. Bah, why didn't I order another chip yesterday? Now I have to wait over Easter if it's toast.
Badwolf wrote: 01 Apr 2021 21:10
We need a list of games that a) run with an 030 and b) use AltRAM.
Yeah, people have been saying the same for years, but nobody ever compiles a list.
I thought just had! :lol:
BW
Re: REV 3 - The beginning
Posted: 01 Apr 2021 22:41
by exxos
Another strange problem with MAPROM running (works fine with just FASTRAM.PRG) ... is when I disabled the I-cache, it locks up on VDI graphics test Pretty much the exact same place every time :shrug:
IMG_6363.JPG
Re: REV 3 - The beginning
Posted: 02 Apr 2021 11:27
by Badwolf
exxos wrote: 01 Apr 2021 22:41
Another strange problem with MAPROM running (works fine with just FASTRAM.PRG) ... is when I disabled the I-cache, it locks up on VDI graphics test Pretty much the exact same place every time :shrug:
Didn't you say you'd hacked MAPROM not to cache the low RAM variables? I don't know much about the VDI, but this feels a bit low RAMmy to me.
BW
Re: REV 3 - The beginning
Posted: 02 Apr 2021 11:35
by exxos
Badwolf wrote: 02 Apr 2021 11:27
Didn't you say you'd hacked MAPROM not to cache the low RAM variables? I don't know much about the VDI, but this feels a bit low RAMmy to me.
Yep.. Floppy doesn't work else.. but it only happens with instruction cache disabled.