Joined
·
1,985 Posts
Originally Posted by Ste Hmm...these don't look as favourable as the Anandtech results. The res is kept low to avoid the bottleneck on the GFX. Would be interested for this site to do the same review at 1280*1024 and above with all the eye candy on. |
Originally Posted by nugget Sorry if this is a double post. http://www.hothardware.com/viewartic...leid=794&cid=1 |
Originally Posted by Ste Hmm...these don't look as favourable as the Anandtech results. The res is kept low to avoid the bottleneck on the GFX. Would be interested for this site to do the same review at 1280*1024 and above with all the eye candy on. |
Originally Posted by HotHardWare.com Conroe's largest lead yet is shown in these low resolution F.E.A.R. demo runs. This test at low res takes the burden off the GPUs and places it more significantly on the CPUs in the two test systems. You're looking at 46% more frame rate from the 2.66GHz Conroe system versus AMD's 2.8GHz overclocked Athlon 64 FX-60. |
Originally Posted by DEMON32 But if they had the bottle neck on the gfx card(s) wouldn't the power all be on them.You wouldn't be able to tell the difference between the 2 cpus. |
Originally Posted by Ropey ![]() 46% more frame rate when the graphics are turned down to remove the graphics subsystem from the equation - good idea. R |
Originally Posted by Ste Exactly and this is with 2 x1900xtxs. The GFX cards still bottleneck the A64s in games, so although Conroe looks way better than the A64s you wouldn't really see the difference in games with current cards. Maybe the next gen will make strides to utilise the CPU more, but to me, this further casts doubts on the validity of the Anandtech results that showed a massive lead for Conroe, even at high res and with high AA, AF etc. Yes, this is great news if you play FEAR at these resolutions. I don't though and I doubt many people really do. |
Originally Posted by Ropey This makes no sense whatsoever as the gameplay at 640 X 480 will remove the graphics GPU subsystem from the equation. Certainly enough to create a valid benchmark (as far as benchmarks go at any rate). This is not about graphics cards, this is about CPU's so let's keep on the track of the new CPU. It's about time to admit that this processor may be far better than the six year old Athlon refreshes. It's not sinful to admit such a new processor is better than the old one and those who maintain that the problem is the graphics subsystem in such benchmarks are using a pitifully useless argument. R |
Originally Posted by Ste I don't think you understand my point. The Conroe looks to be a far superior CPU to the A64 range. However, if the GFX cards bottleneck its performance, in games, then for gaming you won't really see any performance increase until the GFX cards progress enough to fully utilise it. |
Originally Posted by Ste I don't think you understand my point. The Conroe looks to be a far superior CPU to the A64 range. However, if the GFX cards bottleneck its performance, in games, then for gaming you won't really see any performance increase until the GFX cards progress enough to fully utilise it. |
Originally Posted by Ropey That is because your point makes little sense as with dual 7900GTX's and a 40" LCD TV with F.E.A.R. will easily make use of this CPU in 1600 X 1200 and I know this and all you have to do with this graph is extrapolate to 1600 X 1200 as it is at 1024 X 768. Not make use of it? ![]() R |
Originally Posted by Ste But the benchmark in this post uses dual x1900xtxs in Crossfire and needs to lower the res to reduce the bottleneck and show off the CPU in FEAR. |
Originally Posted by Ropey This is the main point. Take the graphics out of the equation. Raw CPU power is what we are talking about here and IS the reason they removed the GPU processing out of the equation. This provides a more equalized platform for raw CPU power. This IS the Intel CPU forum, not the ATI Video Card Forum. Please stay on track. R |