Marketing vs. Reality
In the world of CPUs, market segmentation leads to some strange pricing practices sometimes. Performance of a given chip does not scale linearly with price in many cases, and sometimes, you're paying more for the name than the part.
The main example of this is in Core 2 lineup, which is segmented into Core 2 Duo, Pentium Dual-Core, and Celeron parts. As most techies know, all these chips are Core 2 Duos to begin with; a Pentium or Celeron is made by diking out cores and/or cache that fail the quality control tests. The point being that these chips are Core 2 Duos, just cut down somewhat. What remains still has the same potency clock-for-clock.
The point is, there are other factors to consider besides simply what a chip is called and/or numbered. The most striking example I have seen thus far is the Pentium Dual-Core T4200, a "budget" laptop CPU that nonetheless performs about the same as the mid-high end Core 2 Duo T7500 according to Passmark. What is going on here?
Turns out it's market segmentation at work again. The T4200 is a newer "Penryn" 45nm core, while the T7500 was an older 65nm "Merom" core. Apparently there's so much of a gulf between the two (and the 45nm cores are so much cheaper to produce) that the actual performance matters less than the marketing divisions. Not that this is a bad thing for consumers; on the contrary, this should make a lot of laptop hunters on a budget feel better. The T4200 is more than worthy of the Penryn name, regardless of whether it's called Pentium or Core 2 Duo.
But the reverse is also true, where some companies (Dell, I'm looking at YOU) heavily emphasize brand name over performance. As an example, many of Dell's laptops let you choose between T4200, T6400, and T8100 processors, among others. There is usually a gap of $100 or more between the T6400 and the T8100, but the two chips perform almost exactly the same.
Of course no one would know that unless they did the research, and most people would get suckered in by the "bigger number = better chip" fallacy. Yes, the T8100 has an extra MB of L2 cache, but unless you're doing heavy photo or video editing there's no need for it...and a single MB of L2 does not justify that much of a price gap.
So do your research, and remember: the company is not your friend. You are not a customer. You are a consumer, a "revenue unit." And you are not the target market for CPU makers; the major OEMs and vendors are. Know your needs, know your chips, buy wisely.
The main example of this is in Core 2 lineup, which is segmented into Core 2 Duo, Pentium Dual-Core, and Celeron parts. As most techies know, all these chips are Core 2 Duos to begin with; a Pentium or Celeron is made by diking out cores and/or cache that fail the quality control tests. The point being that these chips are Core 2 Duos, just cut down somewhat. What remains still has the same potency clock-for-clock.
The point is, there are other factors to consider besides simply what a chip is called and/or numbered. The most striking example I have seen thus far is the Pentium Dual-Core T4200, a "budget" laptop CPU that nonetheless performs about the same as the mid-high end Core 2 Duo T7500 according to Passmark. What is going on here?
Turns out it's market segmentation at work again. The T4200 is a newer "Penryn" 45nm core, while the T7500 was an older 65nm "Merom" core. Apparently there's so much of a gulf between the two (and the 45nm cores are so much cheaper to produce) that the actual performance matters less than the marketing divisions. Not that this is a bad thing for consumers; on the contrary, this should make a lot of laptop hunters on a budget feel better. The T4200 is more than worthy of the Penryn name, regardless of whether it's called Pentium or Core 2 Duo.
But the reverse is also true, where some companies (Dell, I'm looking at YOU) heavily emphasize brand name over performance. As an example, many of Dell's laptops let you choose between T4200, T6400, and T8100 processors, among others. There is usually a gap of $100 or more between the T6400 and the T8100, but the two chips perform almost exactly the same.
Of course no one would know that unless they did the research, and most people would get suckered in by the "bigger number = better chip" fallacy. Yes, the T8100 has an extra MB of L2 cache, but unless you're doing heavy photo or video editing there's no need for it...and a single MB of L2 does not justify that much of a price gap.
So do your research, and remember: the company is not your friend. You are not a customer. You are a consumer, a "revenue unit." And you are not the target market for CPU makers; the major OEMs and vendors are. Know your needs, know your chips, buy wisely.