Replies to post #1906 on Advanced Micro Devices Inc (AMD)

[spokeshave]

Tenchu: Re: There is no well-defined system to determine model numbers; it's just whatever number AMD's marketing department feels like slapping on.

On that we agree. Unfortunately, there is no well-defined process that would facilitate determining model numbers. Until recently, the somewhat arbitrary determination has been conservative enough to ensure that the customer got a product that was similar in performance to a similarly labeled Intel product. I am not as convinced with the 3000+. I hope that AMD does not continue to stretch the concept. If they do, they will lose credibility, and the whole QuantiSpeed house of cards will fall.

[Not a Short]

"At least you can't fake a clock speed"

There was an old discussion beaten to death about the P4 having multiple clock regions or somesuch terminology. Some of the P4 operates at full speed some of it operates at a different speed. (Some aregued that there were 1x and 2x portions, others aregued that there were .5 and 1x portions, some went as far as saying there were three sets of speeds 1/2/4 or .5/1/2 depending on the way you wanted to present it)

Do you realy mean to say you can't fake a clock speed? There is no rule that says the MB/BIOS/OS has to report the actual clock speed to you. In fact AMD has shown that the BIOS will happily report a different number than the clock speed.

Irregardless of using the word fake, do you beleive that the Pentium 4 has a single defining clock speed? Does it matter if it has multiple clock regions? If it does is the labeled speed dishonest? Does the performance reletative to the P3 affect that evaluation?

Before anybody things I'm just slinging FUD, let me make my position clear. I don't think the P4 is the worst chip ever (I used to when the 1.3 mhz was compared to PIII and Athlons of the day) but it is obvious that Intel has the deep pockets to make the P4 strategy work. I sometimes think of the Big Dumb Booster comparison (the Russian space program used much cheaper fuel, I'll leave it at that)

So wrapping this whole ramble up, Why is it Quantispeed is bad but you don't complain about other misleading periods in chip history. (I still say any P4 below 1.8 GHz shouldn't have been released with the PIII doing its thing and AMD unable to pick up the slack no matter how far behind INTC got).