Created this since the source is so scattered around this site. All credit to GenTechPC
Fear:
![]()
![]()
G1S( 3gb)
![]()
![]()
![]()
G1S( 2gb)
![]()
![]()
![]()
![]()
Feel free to add more of this type of benchmarks.. We ll see whether if more RAMS make the difference.
-
ViciousXUSMC Master Viking NBR Reviewer
Turbo Cache is whats giving the fps boost from what I can tell.
-
Still, that is a good 10-20 FPS increase.
-
Now, what i wanna check is the benchmark scores with Turbocache disabled.. Does that make a difference here?
-
How much RAM was turbocached? (or can be turbocached)?
-
Well, considering only the first screenshot was even close to the same position, I'm not sure what this means.
If anyone thinks some extra shared system RAM can increase performance 50-100% (which I'm assuming these screenshots are attempting to demonstrate) they're fooling themselves. Maybe if the texture sizes are so large that it's like an increase from 2FPS to 3FPS, but yeah, aside from that...
Perhaps they could run the FEAR benchmark with the different amount of memory? -
It had an effect like that on the F3Sv with 8600M GS too. I think I read that Vista itself hogs around 1GB of memory, the game itself would probably want a few hundred MB, leaving the card very little to draw on. Adding 1GB will give the card more than enough to draw on.
-
ViciousXUSMC Master Viking NBR Reviewer
I agree with chuck in that we need some solid benchmarks to proove anything, a recorded demo or built in performance test so that you know the before and after are exactly the same.
Then you need to test with turbo cache off.
So you need 3 or 4 test:
2gb stock
3gb stock
3gb no turbo cache
4gb stock
With:
> A recorded time demo
> A built in performance test
> Other benchmarks like 3dmark
Then we can have fininte proof rather than just speculation.
Just going into the same general area and taking a screenshot with fraps wont cut it because the fps can change depending on whats in the area (shadows, lighting, computer oponents loaded up into the ram but not on screen).
This is the basis for any real testing and comparisons.
I have been strongly suspicious of these numbers myself since they have come out, while I think turbo cache indeed has to be the reason for the big performance jump the 100% boost some claim seems abit far fetched, unless there was issues with it and its killing performance at only 2gb of ram so the diffrence between 2gb and 3gb is larger due to that. -
You have to ask GenTechPC(who seldom visits the this section) about that, or MrSneis.
I think MrSneis got something like the 3gb magic number thing. Which at 4gb ram it was performing like when it is on 2gb. But at 3gb it just jumped. -
ViciousXUSMC Master Viking NBR Reviewer
Yeah thats the same thing I read, and thats what makes no sense, there is somthing fishy with the turbocache drivers or somthing. No reason for 4gb to do worse than 3gb.
Thats why we really need some hardcore documented benchmarks instead of "hearsay"
edit: here is a nice link to detialed benchmarks of vista running 2gb vs 4gb of ram across an array of games.
http://www.bcchardware.com/index.php?option=com_content&task=view&id=3135&Itemid=40
average about 8% increase of fps. Thats really good considering how cheap ram is and that in the past ram didnt really atribute much to fps but just overall "smoothness" of gameplay because if you didnt have enough ram the pagefile would kick in and you get a pause.
I expect less than this on notebooks because the dont play games to the same settings, but turbocache is obviously whats make the diffrence bringing a moble gpu closer to the levels of a desktop gpu.
Id still go for 4gb of ram and vista x64, thats just what I want and what feels right. There may be issues with it now but when some bios/driver updates have gone thru the works it should be the best performing setup. -
Some 32-bit OSs get slower with 4GB, since they can only address ~3.2GB. With 3GB, they feel safe and unthreatened.
-
What I'd like to know is how Turbocache plays into all of this. I'm only slightly computer literate, but I though Turbocache kinda helps the GPU by unloading some work onto system RAM. Maybe I'm totally wrong. But either way, if Turbocache is helping with 3GB, why would you want to test it w/o it going? What exactly would it prove? Thanks for any clarification.
I know this is a semi-old thread, but with RAM prices kinda low atm (and the lack of many 2GBx2 kits out there being sold), some people may forgo dual channel to go for the 3GB if it possibly helps gaming.
Thanks. -
is turbocache on XP or is it only Vista?
-
with turbo cache off, you would see if the ram difference is helpful for the overall system or the gpu or how much proportionally
-
Can you tell me: is there a FPS boost in games like F.E.A.R. , Crysis, ... with 3GB ram? Because i want to upgrade to 3GB. But i will only upgrade if it HELPS fps boost (to play crysis better, with more FPS)
G1S (3gb RAM) vs G1S (2gb RAM)
Discussion in 'Gaming (Software and Graphics Cards)' started by lunateck, Jun 16, 2007.