This forum uses cookies
This forum makes use of cookies to store your login information if you are registered, and your last visit if you are not. Cookies are small text documents stored on your computer; the cookies set by this forum can only be used on this website and pose no security risk. Cookies on this forum also track the specific topics you have read and when you last read them. Please confirm whether you accept or reject these cookies being set.

A cookie will be stored in your browser regardless of choice to prevent you being asked this question again. You will be able to change your cookie settings at any time using the link in the footer.

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Build Testing With DX12 Reserved For Windows 10
#11
How ;P more fps ?
Reply
#12
(05-24-2015, 08:00 PM)Yagami Light9 Wrote: How ;P more fps ?

It's very unstable now...what are you talking about?
If I helped you, please give the proper reputation.

P.S. sorry, for my terrible english Confused
Reply
#13
That was a question Smile / Yes now unstable may be in future more faster
Reply
#14
Currently it's roughly at the same speed as opengl. Actually it's marginally faster but the difference is not big enough to make it noticeable, and it depends on test.
Performance will likely come later, with modification of RSXThread (currently we copy data several times, which is not efficient...)
Reply
#15
I implemented semaphore in DX12 ; OpenGL makes difficult to use thread that sync with graphic pipeline so I don't think they will be implemented in openGL soon.
Semaphore allow to emulate more accuratly RSX and to take advantage of "undefined behaviors" (ie Cell and RSX reading/writing to same location) to avoid some extra copy.
Hopefully no game bypass semaphore synchronisation between Cell and RSX...

With this optimisation I went from 15 fps with depth write activated to 40 fps in the depth_read.ppu.elf with apparently no regression on other tests.
Reply
#16
Nice speed up and progress @vlj .

Looking forward to the multi texture implementation and this should make some of the commercial games start working .
Reply
#17
Last build 28/05/2015
With Last commit
Quote:d3d12: Send RTTs to cell memory when a semaphore is released
vlj good progress Smile
Now Arkedo Serie 02 Swap show menu and ingame, but crash after 1 or 2 sec
   
   
The Guided Fate Paradox Intro Now color Perfect Big Grin
   
   
Same for Hasbro Family 3
   
   
Upload you Log Here :
PC spec:
Windows 10 PRO X64 Insider 16.232
Amd Ryzen 1700X @3.8 ghz
MSI Core Frozr L
16 go Corsair Vengeance LPX PC4-25600 (3200MHz)
MSI GTX 1080 Gaming X 8G
MSI X370 Gaming Pro Carbon
500 Go SSD Samsung 960 EVO M.2
Reply
#18
Do you know the memory usage of rpcs3 when it crashes?

Visuel Studio 2015 has a tool in the debug tab that display cpu and memory usage history, I use it to track mem leak
Reply
#19
@vij , the depth_read.elf did run pretty quick (full 60 fps on my system) .Not too sure if it is Nvidia card issue , the depth buffer rendered in the bottom right hand corner instead of upper right one.


Attached Files Thumbnail(s)
   
Reply
#20
(05-29-2015, 12:34 AM)vlj Wrote: Do you know the memory usage of rpcs3 when it crashes?

Visuel Studio 2015 has a tool in the debug tab that display cpu and memory usage history, I use it to track mem leak
I d'ont know, I tested in debug, and memory increases, I do not think it is either.
By cons, he points me the line 159 of D3D12Texture.cpp
there must be something that will not, I think.
Upload you Log Here :
PC spec:
Windows 10 PRO X64 Insider 16.232
Amd Ryzen 1700X @3.8 ghz
MSI Core Frozr L
16 go Corsair Vengeance LPX PC4-25600 (3200MHz)
MSI GTX 1080 Gaming X 8G
MSI X370 Gaming Pro Carbon
500 Go SSD Samsung 960 EVO M.2
Reply


Forum Jump:


Users browsing this thread: 11 Guest(s)