Never did I say that the CPU was by any means the driving force behind any large frame rate difference
No...you said.
Generally speaking, which do you think is a better approach for the achievement of higher framerates: AMD?s emphasis on refining the processor core before enhancing clock speeds or Intel?s approach of emphasis on enhancing clock speeds before refining the processor core?
Neither approach is going to affect your framerate in any way.
You insult me by challenging the background of my question, and to this I take serious offence.
There is no reason to be insulted.
**Here is what I feel like...you asked a question about relating 2 unrelated things. it just doesnt make sence.
For example...you could have just asked me: Company A created coffee grinder X to grind whole bean coffee. Its technology is a 10000rpm 2 blade system...it grinds coffee excellently. Company B created coffee grinder Y to grind whole bean coffee. Its technology deptartment believes it grinds coffee better, because it uses a 5 blade system which rotates at 4000rpms. Now...which grinder makes a better tasting coffee??
Ask a relevent question...like which approach to cpu development will lead to more effective processors.
It is assumed by the basis of the question that all factors other than the processor... everything except clock speed, core engineering and L1/L2 cache size is held constant. Two of these factors specifically seem to be perused more consistently by particular manufacturers. Both AMD and Intel are dramatically increasing their L1/L2 cache sizes, while only Intel seems aggressively interested in increasing clock speed and while only AMD seems aggressively interested in adding new levels of capability to their processors (64 bit?).
1)"all other factors other than the cpu is held constant..." impossible. there is not a comparable equilovent MCH north controller chip to create this test with.
Intel uses a quad pumped FSB. AMD uses only a double pumped FBS. The bandwidth is simply too different to provide such a control.
2)"only Intel seems aggressively interested in increasing clock speed and while only AMD seems aggressively interested in adding new levels of capability to their processors (64 bit?)"
Fact: Both company performs core refinements regularly. Example
rescott core. 13 new instructions. better hyperthreading. better branch prediction for piping. So did AMD. Releasing the first 64bit x86 instruction set. Core refinements really only happen every generation of new CPU's. 1st it was mmx, then 3dnow, then SSE and SSE2, P4 pipelining & quad pumped bus, FSB increments, hyperthreading, and AMD dual core. Now its x86 64. I dont see a clear AMD winner here as the only one interested in core revisions.
Fact: The difference between AMD & Intel architecture is the driving force why AMD cant increase its clock speed as much as intel. AMD's cpu logic is much more in depth and advance and
complex than intels. it does more calculations and work per clock cycle, requiring more propogation delay between clock cycles. Intels approach is keep it simple. by simplifying work done per cycle, with no complex logic and less propogation delay, it can cause an instruction to use up two to three clock cycles for completion. therfore pressing speed is not a problem.
FACT: Neither of these CPU improvements approaches affects your video frames per second. It might seem to you that it does because, the items that affect your video are directly tied into the choice of CPU, those being MCH chipset & bus bandwidth...
Dont get caught up in the numbers PR game being played by intel and AMD.
Which tactic is of more merit from a gamer's prospective considering the maintaining of high framerates with ever more demanding games?
-or-
"Which do you think is a better approach for the achievement of higher framerates?"
Now you must understand game programming. Game programs are not cpu intensive applications (comparitively.)
When games run, there are 2 separate program threads or engines running on 2 different processors. There is the game engine which runs on your CPU, and the video engine running on your GPU. They run independely of each other totally, aside from some minimal information exchange. Unless there is a bog down, in your memory/bus bandwidth, and the information exchange is late, there is no effect on your FPS from the CPU.
How can you possibly question the validity of any part of such a simple question?
See above **.
Now in all fairness to your question...I would like to hear some of the ideas bounced around between yourself and you colleages to add to this discussion.
Point me in the direction you want me to speak, and Ill speak in it.
Facebook
Twitter