@All I Will collect benchmarks of my freshly built Sam460Cr with those params:
CPU: 1.1GHZ GFX cards: via PCI-E Riser for vertical setup (coming with the Tower) HDD: Sata SSD 500 GB via SII 3512 SFX: PCI-E SB Audigi FX AmigaOS4 FE Update2 up2date 3D libs : ogles2,warp3dnova,etc.
1. GfxBench results for RadeonRX 560 (Polaris 12) 4gb
But! Should to say, that it's benchmark. When i tested games, they most of time of the same speed for both Radeon RX cards (at least gl4es ones). Mean in this case they pushed to limits because of no DMA in the drivers.
But Spencer game (which written in mind well), have differences in 1920x1080x32xall_maximum_settings:
RadeonRX 570 (Polaris10) - menu 80 fps, in game 45 RadeonRX 560 (Polaris12) - menu 50 fps, in game 37
But this RadeonRX10 is so big beast (with additional power, 2 coolers), and _VERY_ hot. I didn't see that Fans move fast enough even on "HIGH" level, so cards start to be hotter and hotter. Feels like Fans didn't know when to start to move faster (even in High settings).
So for now surely I will choose RadeonRX12 there, even if it slower a bit.
Edited by kas1e on 2021/4/30 16:10:09 Edited by kas1e on 2021/4/30 16:18:15
@Javier Imho our results a bit too low? I just checked the whole site with results, and there some sams which give ~7,000.xx, and a lot of those which give more than ~6,000.xx
@All It is probably worth checking different Radeon RX driver versions... Will do it today. I also have 2 more cards worth testing, will be interesting to see speed change (if any) on the same machine
If memory serves, I think I was getting a higher score with a Radeon HD7770. I think it was around two thousand higher but can't be sure.
Back in months when I test heavy both HD and RX drivers, I find out that RX driver _slower_ in simple small operations. Hans explains that RX needs to do more work behind scenes, and so on small operations, it can be slower.
But in real-life tests, those small operations not very often used (especially in games), and the "big" operations that what is count, and they faster than on RadeonHD. While, because of small tests, overall score can be lower on Radeon RX than with radeonHD.
@Spectre Quote:
May actually have something to do with power settings.
I do all the tests with "HIGH".
Through, what I found, "save" in power prefs just do not work at all! You hit save, reboot, then come back , and blew, the same default settings => low FPS.
Excuse me for being off topic but I have a similar issue with my new Radeon RX570 in my X5000. I get only ~6300 points (Lower than the ~7100 points of my R9 270x) while others score >8000 points. That's with the same library versions.
It is a used card (could have been a mining card even) so I'll check in my PC if the cards bios settings can improve the results. Ultimately I'll reinstall AmigaOS41.
Hmm, thought i'd read a save and reboot worked. Just tried changing to low in power prefs and selecting low. The changed back to high power and selecting use and then save. Ran the test again -
For me, the changes are saved to SYS:Prefs/Env-Archive/A-EON/power.xml . values 0,1 or 2
Damn, as I just install all manually on sam460 from enhancer, I didn't create the "A-EON" directory in the env-archive, so, it just can't save as didn't find the directory. But then, probably it means power prefs need tasty error-window saying "save file can't be created because you happen didn't have A-EON dir in env-arc".
Created directory and it works :)
@geennaam We need to check the same on 2.4.. Will do today on both x5000 and sam460 so we can see if there some general regression in speed and if so, where exactly.
I didn't create the "A-EON" directory in the env-archive, so, it just can't save as didn't find the directory.
That is a limitation of the Application Library (more specifically, its PrefsObjects framework). The library lets you specify a sub-directory for storing your prefs (via the REGAPP_ENVDir tag), but if the path doesn't exist, the library does nothing about it. I think the correct behaviour would be that the library creates the sub-directory for you in such a case.
1. When I save "High power" and reboot then I get a slow result. (6300). 2. When I reboot, open the power pref, press use and run the test again. I get 8260. 3. When I reboot again and run the test I get 6300. When I open the power pref after the first test, press use and run the test again, I get 8260.
Performance is the same for the V2.4 and V2.5 driver.
It makes a huge difference in overal responsiveness of my system.(eg scolling codebench and Odyssey)
Night of the zombies menu difference is 51 vs 81 fps. (+60%)
Will have to wait for the new warp3dnova library to test actual game speed. Now the RX570 doesn't work beyond the menu.
Try just opening the power prefs and then closing them without changing anything. That also seems to work for me.
Can you describe what issue are exactly and how to reproduce step by step (i collect bugs for power tool, just to push them all at one go to the bug-report page).
News feed
OS4 Feedback forum
OS4Depot Feedback forum
Amigabounty forum
OpenAmiga forum
AmiCygnix forum
ABC shell forum
AmiKit forum
Cinnamon Writer forum
CodeBench forum
Digital Universe forum
Dopus 5 forum
E-UAE forum
Gnash forum
Ibrowse forum
JAmiga forum
Odyssey forum
OWB forum
Qt forum
SmartFileSystem forum
Timberwolf forum
TouchDevice forum
TuneNet forum
Unsatisfactory Software forum
Donate
