More CPU benches. >2 cores or bust.
Thankfully thats only at maximum details (aside from view distance). With lesser settings you can get the game playable on a dual core going off some of the benchmarks that have been posted by users.
The big question is, how many of the detail settings are CPU and how many GPU heavy
IGN: Before we get into that, a few questions about general graphics options. Why can't you actually through the graphics menu adjust AA (anti-aliasing), turn off shadows, and why is there a resource usage limitation?
Kevin Hoare: Early on we decided we weren't going to support DirectX 10. We were just going to work on increasing the visual quality with what we had with DirectX 9. If we had DirectX 10 support we could have had the AA in there, but we don't have any. The shadows, at least from the forums and what I'm seeing, I think a lot of people don't understand that the shadow density is only designed for a certain type of light in the game. In the daytime there's a general shadow that covers the entire world. That is hooked up directly to the video mode. It scales automatically with resolution. So the higher the resolution you set the game, the higher your shadow, mirrors, water, reflection, everything automatically scales.
IGN: So they're just all merged into the resolution? Ok so…
Jeronimo Barrera: I wish we could go back in time and name some of the settings a little bit differently because when people are playing it on medium, it's actually, I forget how many times higher resolution it is up from the console version, but it's quite significant.
We had all those options available and we found that people were confused by so many options so we wound up merging them into the video mode
Wait... what? So settings and resolution are not independent? If that's the case Rockstar wins the award for most retarded developers of all time.
The resolution of reflections, shadows and such scale with rendering resolution in pretty much every game, Rockstar isn't doing anything new there.
In the daytime there's a general shadow that covers the entire world. That is hooked up directly to the video mode. It scales automatically with resolution.
==== [AUDIO] ====
-fullspecaudio: Force High-end CPU audio footprint
-minspecaudio: Force low-end CPU audio footprint
==== [GLOBAL] ====
-benchmark: Determines if we run the benchmark immediately
-safemode: Sets the graphics setting to lowest setting
==== [GRAPHICS] ====
-forcer2vb: force r2vb
-frameLimit: number of frames to limit game to
-framelockinwindow: Force framelock to work even in a window (works best with 60Hz monitor refresh)
-fullscreen: Force fullscreen mode
-height: Set height of main render window (default is 600)
-renderquality: Set the render quality of the game. (0-4)
-shadowdensity: Set the shadow density of the game. (0-16)
-texturequality: Set the texture quality of the game. (0-2)
-viewdistance: Set the view distance of the game (0-99)
-detailquality: Set the detail quality of the game. (0-99)
-novblank: Disable wait for vblank (No Vsync)
-norestrictions: Do not limit graphics settings
-width: Set width of main render window (default is 800)
-safemode: Run the graphics in the lowest setting possible
-frameLimit: Limit frame to interval of refresh rate
(ex If refreshrate is 60HZ –frameLimit 1 = Locks down to 60HZ)
-refreshrate: Set the refresh rate of main render window
– Warning - Monitor must support (ex. –refreshrate 60)
-fullscreen: Force fullscreen mode
-windowed: Force windowed mode
-availablevidmem: Set amount of physical Video Memory(ex. –availablevidmem 0.9)
-percentvidmem: Percentage of video memory to make available to GTA
==== [SYSTEM] ====
-noprecache: Disable precache of resources
-nomemrestrict: Disable memory restrictions
I tested these, but only once because I can't find an option for the benchmark to loop n times from the command line."==== [AUDIO] ====
-fullspecaudio: Force High-end CPU audio footprint
-minspecaudio: Force low-end CPU audio footprint"
I wonder if someone with an xfi would benchmark those 2 options (preferably under xp)
LaunchGTAIV.exe -benchmark -renderquality 4 -shadowdensity 4 -texturequality 1 -viewdistance 25 -detailquality 25 -novblank -norestrictions -width 1920 -height 1200 -frameLimit 0 -refreshrate 60 -fullscreen -fullspecaudio
Statistics
Average FPS: 19.41
Duration: 37.41 sec
CPU Usage: 84%
System memory usage: 69%
Video memory usage: 67%
Graphics Settings
Video Mode: 1920 x 1200 (60 Hz)
Texture Quality: Medium
Render Quality: Highest
View Distance: 26
Detail Distance: 26
Hardware
Microsoft Windows XP Professional
Service Pack 3
Video Adapter: NVIDIA GeForce 8800 GTX
Video Driver version: 180.84
Audio Adapter: SB X-Fi Audio [CC00]
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
File ID: Benchmark.cli
LaunchGTAIV.exe -benchmark -renderquality 4 -shadowdensity 4 -texturequality 1 -viewdistance 25 -detailquality 25 -novblank -norestrictions -width 1920 -height 1200 -frameLimit 0 -refreshrate 60 -fullscreen -minspecaudio
Check out the difference in GPU memory usage. That's weird. Memory leak?Statistics
Average FPS: 19.23
Duration: 37.39 sec
CPU Usage: 84%
System memory usage: 68%
Video memory usage: 82%
Graphics Settings
Video Mode: 1920 x 1200 (60 Hz)
Texture Quality: Medium
Render Quality: Highest
View Distance: 26
Detail Distance: 26
Hardware
Microsoft Windows XP Professional
Service Pack 3
Video Adapter: NVIDIA GeForce 8800 GTX
Video Driver version: 180.84
Audio Adapter: SB X-Fi Audio [CC00]
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
File ID: Benchmark.cli
framelimit 30 and minspecaudioStatistics
Average FPS: 18.29
Duration: 37.45 sec
CPU Usage: 86%
System memory usage: 74%
Video memory usage: 89%
Graphics Settings
Video Mode: 1920 x 1200 (60 Hz)
Texture Quality: Medium
Render Quality: Highest
View Distance: 26
Detail Distance: 26
Hardware
Microsoft Windows XP Professional
Service Pack 3
Video Adapter: NVIDIA GeForce 8800 GTX
Video Driver version: 180.84
Audio Adapter: SB X-Fi Audio [CC00]
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
File ID: Benchmark.cli
I have to rerun these minspecaudio tests because I had fullspecaudio in the commandline.txt and it may have overridden.Statistics
Average FPS: 18.72
Duration: 37.24 sec
CPU Usage: 87%
System memory usage: 72%
Video memory usage: 89%
Graphics Settings
Video Mode: 1920 x 1200 (60 Hz)
Texture Quality: Medium
Render Quality: Highest
View Distance: 26
Detail Distance: 26
Hardware
Microsoft Windows XP Professional
Service Pack 3
Video Adapter: NVIDIA GeForce 8800 GTX
Video Driver version: 180.84
Audio Adapter: SB X-Fi Audio [CC00]
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
File ID: Benchmark.cli
Statistics
Average FPS: 18.82
Duration: 37.45 sec
CPU Usage: 85%
System memory usage: 73%
Video memory usage: 100%
Graphics Settings
Video Mode: 1920 x 1200 (60 Hz)
Texture Quality: High
Render Quality: Highest
View Distance: 26
Detail Distance: 26
Hardware
Microsoft Windows XP Professional
Service Pack 3
Video Adapter: NVIDIA GeForce 8800 GTX
Video Driver version: 180.84
Audio Adapter: SB X-Fi Audio [CC00]
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
File ID: Benchmark.cli
I think there's something wrong here.
Okay, awesome.Statistics
Average FPS: 20.97
Duration: 37.25 sec
CPU Usage: 90%
System memory usage: 67%
Video memory usage: 62%
Graphics Settings
Video Mode: 800 x 600 (60 Hz)
Texture Quality: Low
Render Quality: Low
View Distance: 1
Detail Distance: 1
Hardware
Microsoft Windows XP Professional
Service Pack 3
Video Adapter: NVIDIA GeForce 8800 GTX
Video Driver version: 180.84
Audio Adapter: SB X-Fi Audio [CC00]
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
File ID: Benchmark.cli
Statistics
Average FPS: 24.91
Duration: 37.25 sec
CPU Usage: 91%
System memory usage: 67%
Video memory usage: 40%
Graphics Settings
Video Mode: 800 x 600 (60 Hz)
Texture Quality: Low
Render Quality: Low
View Distance: 1
Detail Distance: 1
Hardware
Microsoft Windows XP Professional
Service Pack 3
Video Adapter: NVIDIA GeForce 8800 GTX
Video Driver version: 180.84
Audio Adapter: SB X-Fi Audio [CC00]
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
File ID: Benchmark.cli