i7 930 @ 4ghz HT enabled. PC specs in sig. HD5970 stock (735/4000) QX9650 @ 4Ghz, Gigabyte X48-DS4, 8gb DDR2 1066 5 5 5 15. HD 5970 stock (735/4000) Looks like the Core 2 Quad is a bottleneck for the HD5970. Simon.
To be brutally honest, the conclusion is not valid. 3dmark results are based on different tests, some stressing the CPU and some stressing the GPU. The calculations are averaged in some way, and the result makes up your score. At best, 3dmark is good for comparing gpus to other gpus, but with CPUs the score is almost meaningless. In most games, (GTAIV being the massive exception, and Source to a lesser extent) the difference between one fast CPU and another fast CPU will be minimal, you're talking between three and five fps, tops. So yes, you could say that the C2Q is a bottleneck, but it's only a bottleneck for the 5970 in 3dmark. Apart from that, your scores will differ massively. Try it in a few games. You'll probably be surprised how close the two results are.
I am upgrading from an unoverlocked q6600 to a core i7 930 and I have a 5970 as well so its good to see that this will help me in getting more performance as I thought it would.
The difference wont be that big in most games, i don't really see the point in upgrading to an i7 just for benchies.
Wrong. I have found that running games with 8xAA with the i7 has given me better minimum FPS by a good 10-15%. (Better minimum = Smoother game play Especially in Crysis)
I have never owned a i7 920. But my 930 is Prime stable @ 4ghz. I reckon I could push up to 4.3-4.4ghz. I only paid just over £200 for my i7 930, so the £30 difference is up to you. Simon.
It depends on what your processor is like. Intel has become much better at binning different processors in the last year or so, and so you're less likely to get a brilliant 920. Having said that, Intel's also got better at making Bloomfield cores also, so there are less rubbish processors as well. Personally, I've got my i7 920 up to 4.2GHz with relative ease. Maybe I'd get it a little further if I had a 930, but I'm not sure of that, the limit is likely to be around the same amount.
I never said that the C2Q isn't a bottleneck. I just said that the result you showed, that the score is higher in 3dmark with an i7 doesn't prove that the C2Q is a bottleneck. It's real games you'd want to be testing to try and show that.