An explanation of Morpion Solitair:
http://en.wikipedia.org/wiki/Morpion_solitaire

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

Seems like there are problems with your benchmark.
I get the same speed as an 440ep on the cpu and gfx speed. Strange.

_________________
I'm an antique. Don't light my fuse

@Antique

And your machine is . . . ?

_________________
Raptor BlackBird (awaiting a Miggy like OS)
Amiga 4000 3.9 / Pegasos II 4.1 F.E.

@Antique

Problems are impossible. Results are very stable +- 1point. Maybe refresh screen rate can affect, but it better can explain Chris...

Peg 2 1Ghz Radeon 9200 1GB ram

Test 1 : 81
Test 2 : 17
Test 3 : 419

@eder
I agree. Old radeon cards are faster in 2D operation, Hans talk about it, or has better driver...

Last edited by mrdarek on 03-May-2011 at 08:25 PM.
Last edited by mrdarek on 03-May-2011 at 08:24 PM.

My results:
1. 49
2. Freezes machine.
3. 416

Sam460ex,2gb ram,radeonhd 4850

_________________
I'm an antique. Don't light my fuse

A bit offtopic, but still interesting : as i see from all the latest tests , 2d speed of the radeonhd 4850 drivers are the same as 2d speed of the Radeon9200 and 9250. A bit strange , i was expect much better results through (starting from the PCIe / AGP differences, and ends up with knowing that Radeon9200/9250 are muchmuch worse and older than new RadeonHD4850).

Last edited by kas1e on 03-May-2011 at 09:38 PM.

_________________
Join us to improve dopus5!
zerohero's mirror of os4/os3 crosscompiler suites

Micro with on board Radeon 7000:

78
18
345

_________________
Hate tends to make you look stupid...

AmigaOS 4.1 Classic, Cyberstorm PPC @ 366 MHz, Radeon 9250 128MB PCI:

1) 28
2) 17
3) 179

Runs fine

_________________
AmigaOS 4.x Beta Tester - Classic Amiga enthusiast - http://www.hd-zone.com is my Amiga Blog, check it out!

@Antique
I look into code and not see bugs. It will be more helpfull if you tell me when AmigaOS frezing - on benchmark2 start or befor displayng results page, and if before display results - how is your benchcount value (score=benchcount/100)
@HammerD
good - 28>17 - you are be able play ami-lineapolis at full speed - no need faster machine...

I guess that:

Benchmark 2 is limited by the screen's refresh rate. This seems to be proved by the fact that everyone gets a value of around 17.

Benchmark 1 is limited by the graphics card's bus speed (e.g. PCI or AGP). And so should depend on the motherboard you use, not the graphics card. Sam460 seems to give a not-so-good result, but I guess OS4 beta has not been fully optimised for it yet (and someone getting a freeze seems to suggest it is still a bit buggy).

Benchmark 3 does seem to be strongly related to CPU performance, but maybe memory speed also affects it a bit (I don't know what kind of code it uses). I noticed that CPU MHz divided by two is *very roughly* the same as the benchmark result. Again, the Sam460 seems to be a bit lower result than expected, again probably indicating OS4 (and the firmware) is not fully optimised yet.

Last edited by ChrisH on 04-May-2011 at 12:15 PM.
Last edited by ChrisH on 04-May-2011 at 12:13 PM.

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

@mrdarek

When i press benchmark2 the window gets grey,and mouse locks up. And nothing happens.

_________________
I'm an antique. Don't light my fuse

@Antique
The only difference in benchmark2 is a PortablE command in SetAutoUpdate(FALSE) and later in the code command UpdateAndWaitForScreenRefresh() . This gives a smooth refresh the window. For some reason your graphics card does not accept these commands and does not refreshing the window. This is a graphics driver error. It's very bad news for you - Probably a lot of other games may not run because it's typical technique for displaying smooth graphics, in particular (I'm not sure that) games in window Workbench, full-screen games can work well ... Hans can only help - the author of the graphics driver ....

The up-shot of doing SetAutoUpdate(FALSE) is that all the little graphical updates are combined into one huge update (for when UpdateAndWaitForScreenRefresh() is called). This means that a very large Blit is done once every screen interrupt. Maybe the size of the Blit triggers a bug in Sam460 graphics driver?

Alternatively, SetAutoUpdate(FALSE) also causes a new task to be spawned, which waits on the vertical blanking interrupt. Perhaps WaitTOF() is not properly implemented yet for Sam460 graphics driver?

Last edited by ChrisH on 06-May-2011 at 09:52 AM.

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...