VDP speed curiousity

Page 2/12
1 | | 3 | 4 | 5 | 6 | 7

By hit9918

Prophet (2832)

hit9918's picture

16-02-2014, 00:08

@mars200you,
are you sure this was about MSX2+ vs MSX1.
Wasnt it about MSX2 vs MSX1.

The difference between MSX1 and MSX2 chip is much bigger than between MSX2 and MSX2+.

By mars2000you

Enlighted (5461)

mars2000you's picture

16-02-2014, 00:16

There were also some problems between MSX1 and MSX2, but there were more problems when coming to MSX2+.

By hit9918

Prophet (2832)

hit9918's picture

16-02-2014, 00:26

But the problems of MSX2+ were of the sort "subslot desaster".
While dvik demo makes extreme timing things within one scanline, different topic.

The 9938 versus 9918 is a completely new chip with massively more bandwidth.
The 9958 is a 9938 added horizontal scroll register plus YJK, much more similar.

By mars2000you

Enlighted (5461)

mars2000you's picture

16-02-2014, 01:19

You're right : the skipped parts on the Utopia demo are skipped also on MSX2, these parts have interlaced pictures in screen 2, using more than 100 colors. However, on the remaining parts, you can notice some glitches when running on MSX2+ or turboR, I guess they come from the difference of VDP speed for some instructions.

By hit9918

Prophet (2832)

hit9918's picture

16-02-2014, 01:42

problem is, both this mystery benchmark and MSX1 demo do port 98,
and that is all dictated by z80 timing, no VDP status register checks.

but when we are talking extremely timed MSX1 demo. there is things like
horizontal position of vblank interrupt
hotizontal position of changes in 5th sprite status
(i.e. special demo actualy does mess much with status register)

and that maybe is a bit different timing on 9938.
and who knows, maybe something again a bit different on 9958. its horizontal scroll register may have made some cycles difference.

By hit9918

Prophet (2832)

hit9918's picture

16-02-2014, 01:50

but for how this benchmark can go 5% different on MSX2+ I got no explanation.

the explanation CANNOT be "VDP speed".
because z80 OUT business is always the same speed.
the causes can only be "some other category".

By Daemos

Paragon (1669)

Daemos's picture

16-02-2014, 01:55

The difference is not only noticable with this benchmark other benchmark tools show similar results. I think the engine on the msx2+ and turbo-R insert some extra wait states towards the CPU when outing.

By hit9918

Prophet (2832)

hit9918's picture

16-02-2014, 02:06

mhm as numbers are different on bluemsx, too, emulator writers long know about it?

so far I only heard about such things in R800 mode.
VDP brake and some RAM refresh hiccups.
but the exaggerated VDP brake makes much more than 5%.

And then I heard about MSX2+ that can do 5.4Mhz.
Kinda Turbo light.
Is it maybe those boards that make a difference.

By hit9918

Prophet (2832)

hit9918's picture

16-02-2014, 02:20

Can TurboR too run z80 in 5.4Mhz?
Is that faster vram transfer than R800!?
Some findings here:
http://www.msx.org/forum/msx-talk/hardware/need-help-t9769-p...

Quote:

T9769B and T9769C are pin-compatible. The Z80 core in T9769B can only work at standard speed (1x - ~3.58MHz), while T9769C Z80 core can be soft-switched to "turbo" mode (1.5x - ~5.4Mhz). All Panasonic MSX2+ computers are equiped with T9769C (A1FX, WX and WSX), while all other MSX2+ models, both from Sanyo and Sony, are using T9769B

Quote:

For that speculation to be true, the turboR's Z80 must be switchable to 5.38MHz too.
Who will test this?

By NYYRIKKI

Enlighted (5211)

NYYRIKKI's picture

16-02-2014, 04:52

Quote:

For that speculation to be true, the turboR's Z80 must be switchable to 5.38MHz too.
Who will test this?

That would be very nice... but... No...

Page 2/12
1 | | 3 | 4 | 5 | 6 | 7
My MSX profile