RPGW1ZaRD
December 8th, 2006, 17:28
I've known it since launch of this new version but I haven't really showed you before and I've seen lots of threads regarding "how will this comp run X game" or "what should I upgrade" etc. so that's why posted this after all.
I was using TLB build when taking all the screenshots below as only that is working with GSdx9 in v0.9.2 with MTGS & DC enabled.
Computer Setup
Opteron 165 @ 2.8GHz
nVidia 6800GT 256MB @ 440/1140MHz
OCZ Gold VX PC4000 2x512MB @ 2-2-2-5-1T
ZeroGS settings: 1024x768, PS3.0
GSdx9 settings: Windowed, PS3.0, LTF
PEOPs settings: fastest settings and everything unchecked
CPU: settings: EERec, VU1Rec, VU0Rec, MTGS & DC mode, normal
PCSX2 v0.9.1
-------------
GSdx9 0.9.0 (http://img169.imageshack.us/my.php?image=testre4091gsdx9nn9.jpg) & ZeroGS 0.95.2 (http://img171.imageshack.us/my.php?image=testre4091zerogsvg5.jpg)
PCSX2 v0.9.2
-------------
GSdx9 0.9.0 (http://img454.imageshack.us/my.php?image=testre4092gsdx9gq7.jpg) & ZeroGS 0.96.0 (http://img365.imageshack.us/my.php?image=testre4092zerogsxk9.jpg)
As you see CPU usage differs quite a lot here between 0.9.2 but I don't really know why GSdx9 speed in both PCSX2 v0.9.1 and v0.9.2 remains the same altho cpu usage changes from 80 to 100%. But at least I can say ZeroGS 0.96.0 has become less GPU dependant in the current release, here we see if we compare the FRAPS value which is more of an average value it differs even 43 vs 26 FPS which is a nice ~65% difference and CPU usage went up from 75-77% to a steady 100%. This doesn't apply only to Resident Evil 4 but also a couple of other games which are more demanding, such as SmackDown Vs Raw 2k6 & 2k7 mostly when the wrestlers walk in, Tekken 4, Grandia III during fights of what I've been able to observe myself.
So I just wanted to say I think zerofrog has made some optimizations for v0.96.0 making it more efficient somehow so you won't need as good gfx card for v0.9.2 as for v0.9.1 and this hasn't ever been meantioned before so that's why I thought it could be meantioned.
I was using TLB build when taking all the screenshots below as only that is working with GSdx9 in v0.9.2 with MTGS & DC enabled.
Computer Setup
Opteron 165 @ 2.8GHz
nVidia 6800GT 256MB @ 440/1140MHz
OCZ Gold VX PC4000 2x512MB @ 2-2-2-5-1T
ZeroGS settings: 1024x768, PS3.0
GSdx9 settings: Windowed, PS3.0, LTF
PEOPs settings: fastest settings and everything unchecked
CPU: settings: EERec, VU1Rec, VU0Rec, MTGS & DC mode, normal
PCSX2 v0.9.1
-------------
GSdx9 0.9.0 (http://img169.imageshack.us/my.php?image=testre4091gsdx9nn9.jpg) & ZeroGS 0.95.2 (http://img171.imageshack.us/my.php?image=testre4091zerogsvg5.jpg)
PCSX2 v0.9.2
-------------
GSdx9 0.9.0 (http://img454.imageshack.us/my.php?image=testre4092gsdx9gq7.jpg) & ZeroGS 0.96.0 (http://img365.imageshack.us/my.php?image=testre4092zerogsxk9.jpg)
As you see CPU usage differs quite a lot here between 0.9.2 but I don't really know why GSdx9 speed in both PCSX2 v0.9.1 and v0.9.2 remains the same altho cpu usage changes from 80 to 100%. But at least I can say ZeroGS 0.96.0 has become less GPU dependant in the current release, here we see if we compare the FRAPS value which is more of an average value it differs even 43 vs 26 FPS which is a nice ~65% difference and CPU usage went up from 75-77% to a steady 100%. This doesn't apply only to Resident Evil 4 but also a couple of other games which are more demanding, such as SmackDown Vs Raw 2k6 & 2k7 mostly when the wrestlers walk in, Tekken 4, Grandia III during fights of what I've been able to observe myself.
So I just wanted to say I think zerofrog has made some optimizations for v0.96.0 making it more efficient somehow so you won't need as good gfx card for v0.9.2 as for v0.9.1 and this hasn't ever been meantioned before so that's why I thought it could be meantioned.