Joe, Re: "Thanks for the response. I think we can get to the bottom of this."
You're welcome, and I apologize for the unkind words. I simply don't agree with the way some people are handling this. Jumping to conclusions based on an incomplete understanding is something I'd expect from Dan3, but not intelligent people such as yourself.
If you want to continue your analysis of the Pentium 4, at least make it more meaningful than just trying to prove if Intel's megahertz label is justified or not. Calling a 2GHz Pentium 4 2GHz may not be 100% technical, since there are multiple clocks in the system, but we have to take Intel's word for it that a majority of units inside the CPU run at the nominal clock.
Your previous post tried claiming that the decoder speed would be better suited towards labeling the chip, which I'm sure is why you'd like to see it proven that it really runs at 1GHz. However, a good scientist doesn't work a result towards a previously conceived goal. You are forcing the conclusion, and that's what has made me frustrated.
I, too, would like to know the truth, but I don't want the truth perverted by the AMDroids to satisfy their selfish goals. They like AMD so much they want to see the competition fail. Consequently, they want the Pentium 4 to fail, or at least taken off the pedistal upon which it has been placed by the industry. They think it's unfair that a processor can be labeled to be faster than it really is.
However, instead of attacking the problem, they are over-dramatising certain facets in order to brute force a confrontation. They want people like me and Paul to see an obvious defect in the design. They want to prove that there is a definite inconsistency with the implementation.
Finding such an inconsistency or defect, though, may prove quite difficult. You see, the engineers of the Pentium 4 weren't idiots, and Kap is trying to attack the problem under the assumption that they are. Some of the brightest people in the industry work for Intel, and although not every one has a huge bearing on the design, I'm sure that the team as a whole would not be so silly as to overlook the obvious.
Therefore, whatever performance problems that the Pentium 4 has, you are not going to find it with amature test programs. We can only assume that the few people intimately close with the design know how to improve it over time. Northwood will be interesting to look at, and see if any obvious changes have been made. I am sure that at least a few things were changed at the micro-architectural level - and not just the added L2 cache.
So rerun the program if it makes anything more clear to you. Alter it a little bit, and see if you can make some conclusions. However, I would not make any more conclusions like you did in your previous post. All of a sudden claiming that a 2GHz Pentium 4 should be labeled at 1GHz is unreasonable to say the least.
wbmw |
