Thread: Gtx 570 Phantom bp perfs

I have no performance gain expected. Crysis, the game oar systematically from the moment I turn on AA or I put myself in 1920x1080. (I have a screen 23 ").

I tested Crysis because I thought the performance limitations on the demo Crysis2.

A friend who has an ati 4870 have better performance (fps does not drop below 30) than me who have an average of 42 FPS with falls was 20 ...

Here is my config:
Win7 64bit
Asus P5K
Core2Duo E8400 3.0ghz
Gainward gtx570 phantom
4GB dual channel kingston (1080mhz)
WD Raptor.

I tested the demo and I Metro2033:
Frames, Time (ms), Min, Max, Avg
3488, 60000, 37, 85, 58,133

Starcraft2 on, the framerate is in the 50 medium-counting phases big fight. So I do not know what to think, you know utilities that test the GC by comparing with the results of factory outlet?

Okey crash test, that's what I fear. I trusted my local dealer when I changed my case (power supply cable from my previous method was too short), I would not! I changed right away for Antec NeoPower 650blue have a friend on a PC and I'm telling you the results.

Well, I put the Antec NeoPower 650 Blue. It changes nothing performance level but I am more quiet! About the processor, I recently wanted to change but seeing the tiny gap between the i7 and I dropped my model.

He got a Core2Duo E8500 not OC. Frankly I longuemte ent Tate the process but since the socket must be changed RAM CM Proc .. and given the evidence I have decided to stay in this config.The purchase came after the CG so I'm not asked, that the process flange est_ce GC?

I went up to 3.9 but I have many instabilities, even with 3.4 I have some freezes at times (no crashes), I tested with a Vcore auto, lowered ram frequency. I have not yet changed the RAM cycles (5-5, etc.).

Actually I just started documenting myself but given the instability I'm not too excited and I feel that the quality of a CB parameters depend quite invariable, it's not even copy and paste if I return the settings of a guy who did the same components as me.