Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Waiting for Gabest's revival (DX 12 Support)
#31
(01-25-2015, 06:42 PM)rama Wrote: If DX12 does the same things as DX10/11 did, only at reduced CPU costs, then it would help PCSX2 speed, probably a good deal even.

This is probably also the worst case scenario

Sponsored links

#32
(01-26-2015, 09:49 AM)Serial Hacker Wrote: This is probably also the worst case scenario
Worst case scenario ? I think you are overestimating DX12, sure it has almost doubled the performance in pc games but, emulation is entirely a different case. DX12 may/may not function same as it's predecessors so, we cant be sure of it yet.
We're supposed to be working as a team, if we aren't helping and suggesting things to each other, we aren't working as a team.
- Refraction
#33
(01-26-2015, 10:25 AM)ssakash Wrote: Worst case scenario ? I think you are overestimating DX12, sure it has almost doubled the performance in pc games but, emulation is entirely a different case. DX12 may/may not function same as it's predecessors so, we cant be sure of it yet.

the worst case scenario is indeed only lowering the CPU overhead and not offering any new functions or monumental changes
#34
(01-26-2015, 01:59 PM)Serial Hacker Wrote: the worst case scenario is indeed only lowering the CPU overhead and not offering any new functions or monumental changes
that can't be considered as the worst case scenario Tongue , we have absolutely no proof that, it will lower cpu overhead in case of emulation. so, lets not come to a conclusion with only assumptions.
We're supposed to be working as a team, if we aren't helping and suggesting things to each other, we aren't working as a team.
- Refraction
#35
(01-26-2015, 02:32 PM)ssakash Wrote: that can't be considered as the worst case scenario Tongue , we have absolutely no proof that, it will lower cpu overhead in case of emulation. so, lets not come to a conclusion with only assumptions.

i don't think the new set of instruction would behave differently under different atmospheres

1+1=2 everywhere.
#36
(01-26-2015, 06:41 PM)Serial Hacker Wrote: i don't think the new set of instruction would behave differently under different atmospheres
Any points to back it up ?
We're supposed to be working as a team, if we aren't helping and suggesting things to each other, we aren't working as a team.
- Refraction
#37
Why do you think it will be different? PCSX2 is just a program. Games are just programs. What makes PCSX2 special as far as DirectX is concerned. DirectX isn't aware of the type of the program using it. It's not like DirectX will say "Oh, you're PCSX2. In that case, the CPU's workload won't be reduced."
#38
i have to agree with xemnas, OGL performs the same in Windows and Linux.
#39
The point is that DX12/Mantle aren't magical fix-it's like people think they are. On PC games, they reduce CPU overhead to allow the GPU to do as much work as it's supposed to and more efficiently at that. It does this by removing much of the bloat of D3D and OpenGL that's been there for ages. Less cycles wasted on a badly-written API so you can squeeze out many more draw calls than you usually could, for example. Inherently proper multi-threading to allow for even less of a CPU bottleneck. But you couldn't get this last bit in PCSX2 unless the entire emu got re-written. Otherwise it's impossible to pin down the extent of the performance increase something like D3D12 would provide to PCSX2. It would probably be a noticeable boost, but I don't think it'd be something mind-blowing or anything. We'd have to see it in action first.
#40
with now over 80 million Windows 10 users, any light being shown over this ?

Gabest plz ;_;




Users browsing this thread: 1 Guest(s)