The Spin off Smoothness Thread

[freeloader]

This is strictly my own personal opinion.

I've had a Q6600@3.5ghz on a DFI board for about a year now. I also have a Phenom 9850@3.0ghz on an Asus M3A79-T mobo. Contrary to what anyone tries to tell me, the Phenom rig just seems to run better/smoother. Maybe it's a placebo effect, maybe it isn't. I can only go by what my instinct is telling me. I use the 9850 as my daily rig while the Q6600 sits folding away in my basement now.

[Final8ty]

Simple.. Intel is like running a game 100+ fps with V-Sync Off while AMD is like running a game with V-Sync On 60fps.

When i play games @ 100+fps V-Sync off it seems jerky to me & that's why i play all my games with V-Sync on even tho its only 60fps its smoother.

[keithlm]

I don't think you can run a game benchmark on the two systems and compare the min, max, and average FPS. That won't really show anything since the systems are not really stressed in that situation. You need to add stress.

If you do the same benchmark with OTHER things taxing and stressing the systems you should expect a decrease in both systems. The question would be whether or not the decrease is equal for both systems.

If the decrease is equal for both systems then the "smoothness" factor is debunked. But if the performance drops more suddenly on one system and is more gradual on the other system... then that could prove not only the existence of the smoothness factor but could also be measured.

AND who is to say that it won't be Intel with the more gradual decrease and AMD dropping suddenly? (Assuming they are not equal.)

ANWAY: The problem is determining some kind of various tasks that can be run while doing the other benchmark that will equally stress various resources in the system. I have seen a few people try gaming while running a virus scanner and that wasn't enough to really affect the system enough to make a determination. We need to determine various tasks that will eat away at certain resources until we get results or completely debunk the concept.