configuration (hardware / software) for wii and ps2 games

Retroboy · 2019-10-05T17:54:41+00:00

Batocera 5.23 x64 Intel core i5-2500 3.3Ghz 4GB Ram (Intel HD Graphics 2000) no configuration changes Good morning, everyone! I just installed Batocera, and...

Tomsen

PaulMcClernan I'm pretty sure a better GPU will help a lot.

Tomsen

joinski If tested PS2 with a ATI hd 5750 and a ATI hd 7750, if your CPU is powerful enough I think you can play about anything and these cards are very cheap to buy. ATI hd 7750 doesn't even need extra PC- express connectors (12v).

joinski

Tomsen
nice to know! 🙂 so HD 7750 is a cheap alternative

Dragu

I played through Wii vers. Golden Eye, and finished the game with the J5005Itx board.
The 2,8Ghz (rated 1,5GHz, this processor mostly run at turbo speed) are still ok, but not perfect with the internal Gpu. Now, try to find an performant graphic card for the board Pcie 2.0 connector....

joinski

Dragu
i guess when you use a graphics card like Radeon HD7750 or GeForce GT 1030, then your CPU will be the bottleneck 😆

Dragu

joinski
The only card to fit in the slot of the J5005itx board is the Zotac GeForce GT 710 PCIe x1.
I think I have just the place, I expect better performance as with the integrated GPU605.
What do you think. I dont want to go to the edge of most performant machine with a noisless
10Watt CPU. Just try an upgrade.

joinski

Dragu

You could give the GT 710 a try, according to 3DMark it has about 63% better performance, than the Intel HD 605 Graphics of the J5005 😉
https://www.videocardbenchmark.net/compare/Intel-UHD-605-vs-GeForce-GT-710/3861vs2910

Could you, as an alternative, not use something like this:
https://www.conrad.de/de/p/renkforce-pcie-x1-x16-adapter-riser-card-bitcoin-1616188.html
https://www.amazon.de/Silverbead-Extender-Verl%C3%A4ngerung-Adapterkarte-Erweiterungskarte/dp/B07SDRMTLJ/ref=sr_1_3?keywords=pcie+x1+x16+adapter&link_code=qs&qid=1571155446&sourceid=Mozilla-search&sr=8-3
togehter with a standard pcie x16 graphics card?

Dragu

joinski
Thank you ! I just found out that there is newly also an Zotac GT 730 card with the same connector.
I will try to find this one. https://www.zotac.com/us/product/graphics_card/gt-730-pcie-x1

Dragu

joinski
Something funny. As you wrote and as I read, the gtx710 should be about 63% more performant as the internal UHD605.
So I give it a try today. And big surprise, with the pcbench dos graphic test (only test existing for Batocera, Recalbox)
I arrived before inserting the card to 36,4 (always depending of the buildroot version) and with the graphic card to 31,1. Remembering that the PI3 had an index of 5,1. So seems that the Intel internal gpu is more speed on a linux system. Even a poor UHD605. So, I think my next upgrade will be the J5040 (3,3GHz) 10W coming out in 2020. If Asrock make a matching itx board again. 🙂 What is clear, concerning our gaming software without special drivers you cannot rely on passmark values.

Tomsen

Dragu is IT all to save power? OR do you Just want A small build?

Dragu

Tomsen
It was the choice to replace a PI3 in a Bartop, with limited place. And I found that for the heat issue inside, I would need a CPU about 3GHz but passive cooler. As such I found the Asrock J5005-Itx and I am really satisfied with the choice, as for most of games that I am interested, it was a good choice into a small space with as 7x higher speed as the Raspberry. Later, as in this summer it was crazy hot, temperature climbed about 70°. So I also put an 14cm fan turning just with 240 rpm. Always no noise but 45° under all conditions. So some place for a passive graphic card. 🙂
https://batocera-linux.xorhub.com/forum/d/1743-j5005-vs-pi3b-processor-power

Tomsen

Dragu aaah lol now I understand. Well I had a Pi3 and was unable to play N64, so I bought a Odroid Xu4 from korea and then I was able to play N64. Then I saw a youtube movie from ETA prima showing a simpel computer that was quit cheap emulating much better, so I decided to go for a ''small factor form HP'' so I was thinking that would also fit in your bartop, but im not sure it does, and offcourse it is not a 10 watt cpu, anyway i also upgraded that pc to a i3-3220 at a max of 55 watt, with a ATI 7750 (no fan). I sold the other things. Down the road I have spended so much money, but I am able to play about anything. (just a few games lag). Your bartop looks great by the way.

Dragu

Tomsen
It also took some time for the making, you can find more about here (in german)
https://www.antik-automaten.de/wbb2/thread.php?threadid=2206
Some illuminated buttons are home made, rainbow buttons, and the 2 player buttons with own fader controller,
the program is also there. Old kid playing 😉

Tomsen

Dragu mein Deutsche ist nicht so gut;-). Buy i can Read it. Very inpressive! Nice stickers, the whole bartop looks great.

joinski

Dragu

are you sure, you can rely on pcbench as a valid performance benchmark for gpu performance?
i know you can not 100% rely on passmark values (i think you can not rely on any benchmark-software, because every of them works different).
so the best would be, to do all different benchmarks out there and then you get at least a "feeling" or an approximately estimated value of performance.

but the very best would be, to compare the performance within the actual systems/games (dreamcast, 3ds, gamecube, ps2 etc.)

Dragu

joinski
I think that all this tests are based on Windows. So, in the bios of this card that I tested supports directx 12.
No support in Linux. But concerning OpenGl 4.5, it is supported by UHD605 and the tested Zotac card.
the bench program use no drivers at all, just pure mathematics. What you get is a combination of your cpu
speed together with the video memory speed. And here seems to me that in this case the chip is more speed
then the transfer over the card. It would be really interesting to have a benchmark program for our software.
As I remember Substring was planning something like this at the end of his Recalbox time.

joinski

Dragu

yeah, a benchmark program for batocera would be cool.
but as i already mentionend, you could also just activate fps-counter and test ps2, 3ds, gamecube, ...etc. games instead.
i did this a few month ago with a beta batocera 5.21, see here:
https://docs.google.com/spreadsheets/d/1K-HDv3SXqWj-aoniuAJnFzsRNsg8HyHv55C0RJgFgrY/edit#gid=0

as you can see in line 14 and line 15, with intel hd 4600 graphics (integrated gpu of the intel i7) the pcbench score was less then with the same processor together with a nvidia GTX 750 Ti.
but in games like Mario Kart Double Dash for gamecube i could raise the resolution to 6x native and the game was still running in full speed with 60fps (with intel hd 4600 i could only set 2x native resolution. with 3x native the game slowed down and was not running in constant 60fps anymore).

Tomsen

joinski what does 2x mean? and why is FPS score not mentioned? I don't get the scores, that could be because I just don't know where to look at? passmark single an multi score is what I understand, but not how the games ran?

Dragu

joinski
I agree with @Tomsen. Nice job from you. There are some benchmarks ready for buildroot. This could be an opportunity. But like the echo from devs approch 0, I doubt the time investment and honestly would prefer if older issues were resolved, before beginning something new. We were going from 5.20 on now through nothing more as just some beta versions .21 .22. 23. Better to slow down the speed of new versions and first fix the problems in the versions before. I think that exactly this made loose some users to Batocera. Its asthonishing calm here.

joinski

Tomsen

0 means the game did either not run, or run slowly (not fullspeed)
1x means that the games run fullspeed (30 or 60fps) in native resolution.
2x means the game did run fullspeed at 2x native resolution
3x = 3x native resolution
and so on

for me a system/game is only playable, when it runs in 100% fullspeed (60fps or 30fps, depending what the native fps of the system/game is), because i hate micro-stutters and lags

Tomsen

joinski Thanks, nice spreadsheet.

« Previous Page Next Page »