4ghz i5 750... lower Vantage GPU score?

Last night I was fooling around with overclocking. I finally decided to forgo Turbo Boost and see what sort of OC I can get. No turbo has some advantages, the power usage in HWMonitor went from 108W to 95W. The temps dropped too for the same speed. My Turbo Boost OC was at 175 base clock, for 3.68ghz to 4.2ghz. For the tests, my CF 5850s GPUs were at 960/1200.

So here's my initial results with the 175 base clock + turbo boost set up:
VantagePerformance-PPUOFF.png


Pretty darn good! And this is with PPU disabled mind you, as with my PhysX card I got something like 30k CPU score lol. So 16k is accurate.

I took a screenshot but I didn't upload it as it was getting late, so you'll have to take my word on this, but after the 4ghz OC, I scored 18k CPU which is great, a 2k increase. But, the GPU scored was actually lower by about 1k I think. I noticed it showed in the side panel data there, the FPS were lower in both GPU tests as well as a couple feature tests. So I'm trying to figure out why?

Does Vantage run single core on the GPU tests? I did have the windows performance monitor open, it did show around 25% usage for those parts but it was spread across all CPU cores according to the graphs. Still, that's the only reasonable conclusion I can come to unless somebody out there can give me some new insight. Basically, if it's running single thread GPU then it was only going 4ghz, and therefore bottlenecked compared to 4.2ghz with the turbo OC. One thing I'm going to do is run some gaming benchmarks and see if those are performing alright.

The only other thing I can think of is that the RAM speed changed from 1400 CL7 to 1600 CL8. I am going to work on maximizing the RAM speed but I don't think that should make too much of a difference... Or maybe it does?
 
Well, I did a rerun immediatly after and got pretty much the exact same scores. I guess I can put in the time and try the test at, say, 3.8ghz and 3.6ghz and see if my gpu score changes each time. Or try 21x200 for 4.2ghz and (assuming I can get it stable) see if I'm back up to the previous GPU score...
Still, pretty crazy that a 4ghz i5 750 might bottleneck Vantage... Ohh... maybe an easier test would be to run it on High settings since that boosts the resolution and all that. I think I have a High run from before.

Also, may not matter, but I recall now that during the first GPU test, there was a significant slow down at the part where the crate drops into the water, which I don't really recall happening before.
 


No they're the reference design Sapphire cards (Dirt 2 edition). Yea, they needed more voltage I think I'm running 1.175V.

And, well... yeah... lol it's a pretty good score :D
 
Nope haven't changed it. But yesterday I did take it right off to have a look, I wanted to see if I could put some Zalman VRM heat sinks on it but there's no room. I might look into aftermarket cooling but eh, they're running pretty well. And yeah it can get a little loud when it's really going but if I'm gaming, I don't notice usually.
 
Alright Wolfram, lol thanks you and your telling of your reference cooling use! I went back to GPU Ocing, and wow. I was surprised when i saw that i could actually push the reference pretty far! I OC'd it stable to 900/1205 and no voltage changes :D Runs awesome, i did reach 950/1205 but it was unstable and so YEA! Ima test out some games and thank so MUCH!