Replies to post #29092 on Advanced Micro Devices Inc (AMD)

[Tenchu]

C, it underscores my criticism I made of the model number scheme when AMD introduced it: It's all arbitrary.

Then again, it reminds me of the chipset numbering scheme. Remember the 820 RDRAM chipset and how the 815 PC133 chipset (as well as a 440BX overclocked to 133 MHz) outperformed it?

Tenchu

[wbmw]

Constantine, the system has some arbitrary aspects, but it's not hard to figure out. 300 series -> 500 series is moving from Celeron to Pentium, so the added value is in performance. You can't compare the digits, but you know that 500s are generally faster than 300s. The numbers inside each family tend to equate to similar performance, but only within the family, not across both families. This does not seem to be the case with the 500 series -> 700 series. The inner digits to seem to match one another relatively well, but the added value comes in the form of lower power (i.e. mobility). Thus, a 745 may not have stronger performance over a 550, but it might still be worth paying more for (it's the luxury model ;-). It also makes it easier for people to choose 530 over 330. The 530 will be more bang for the buck, while the 330 is the entry level offering.

[sgolds]

wmbz, I'm kind of surprised over the negative reaction I see here in several posters with regard to the Intel model numbering scheme. We have long complained about the Mhz Myth, and how Intel should get away from that scheme.

Neither the Opteron model scheme or the Intel model scheme is a TPI number. They are simple model numbers with performance comparable within a class but not easily between classes (what is the relative performance of a 4-way 842 vs. a 2-way 242 system?). What's the deal with the second digit, X4X, anyhow? What does it mean? It seems to indicate Opteron, just as the first digit in the Intel scheme indicates the family of processors.

I am glad Intel is finally admitting the bankruptcy of the Mhz Myth.