- Oct 13, 1999
- 22,377
- 2
- 81
Regardless of what zealots here and elsewhere say, I do like reading Tom's Hardware articles. Seems like every hardware site has their quirks and biases. Information is information, and the more you have (to a point) the better. Sometimes Tom's has unique information or ideas not found (yet or at all) elsewhere.
Two pet peeves of mine are bad editing (grammar/spelling, not that I'm perfect but some shadier sites are atrocious) and inaccuracies. I feel that online publications (any web site, not just Tom's) should be held to the same standards as print publications. I do cut some slack since sometimes the author obviously does not natively speak English. I find great interest in seeing HardOCP articles starting with the line "Grammatical & Spelling Editor: Timothy Daniel." I don't think it deserves that much attention, but I do appreciate the fact that at least some effort is made in that direction.
With that being said, I read Tom's Hardware's article "The Mother of All CPU Charts Part 1" with much anticipation. A trip down Memory Lane, so to speak.
Then, came the inaccuracies and omissions.
Some are pretty small, making me doubt my memory.
Others... ticked me off a tiny bit.
Example: "using an adapter, the PPGA CPUs could also be used in Slot 1; for the FC-PGA version there was no adapter." WTF? The PPGA, FC-PGA and FC-PGA2 chips all had slot 1 adaptors available at one time or another.
"AMD set the example - Intel followed it: the introduction of a new model numbering scheme that ultimately obscures the reference to the clock speed." This was mentioned on the Socket 775 section. "AMD entered its model numbering in the struggle for performance ascendancy, which cause a good deal of confusion. The first XP CPU was 1800+, but only clocked at 1,533 MHz." This was mentioned in Part 2 of the charts, in the socket A section. What about Cyrix's earlier "PR" efforts? They were so blatant that some chips saw their PR rating magically rise over a period of time, even if the chips themselves were no different (though you needed a BIOS flash to "magically" get the higher speed). What about AMD's earlier attempts with the K5 chip (itself a footnote in Tom's article)? Speaking of the K5 chip, what about the fact that they were "like" the Athlon chips in that originally they were sold/marketed at their actual MHz, and after some optimizations they were sold/marketed using a "PR" speed? What about AMD's original? I'm speaking of the chip that was in reality a 486 core (perhaps with optimized/increased cache) running at 133MHz. Some people called it a 486DX4-133. More "officially" it was a 5x86-133 P75. Yup, AMD was calling a 486 chip on steroids a Pentium 75.
Where are all the others? The IBM/Cyrix chips? The IDT/Winchip? The NexGen 586 chip? These should have their place in history.
Comments?
Two pet peeves of mine are bad editing (grammar/spelling, not that I'm perfect but some shadier sites are atrocious) and inaccuracies. I feel that online publications (any web site, not just Tom's) should be held to the same standards as print publications. I do cut some slack since sometimes the author obviously does not natively speak English. I find great interest in seeing HardOCP articles starting with the line "Grammatical & Spelling Editor: Timothy Daniel." I don't think it deserves that much attention, but I do appreciate the fact that at least some effort is made in that direction.
With that being said, I read Tom's Hardware's article "The Mother of All CPU Charts Part 1" with much anticipation. A trip down Memory Lane, so to speak.
Then, came the inaccuracies and omissions.
Some are pretty small, making me doubt my memory.
Others... ticked me off a tiny bit.
Example: "using an adapter, the PPGA CPUs could also be used in Slot 1; for the FC-PGA version there was no adapter." WTF? The PPGA, FC-PGA and FC-PGA2 chips all had slot 1 adaptors available at one time or another.
"AMD set the example - Intel followed it: the introduction of a new model numbering scheme that ultimately obscures the reference to the clock speed." This was mentioned on the Socket 775 section. "AMD entered its model numbering in the struggle for performance ascendancy, which cause a good deal of confusion. The first XP CPU was 1800+, but only clocked at 1,533 MHz." This was mentioned in Part 2 of the charts, in the socket A section. What about Cyrix's earlier "PR" efforts? They were so blatant that some chips saw their PR rating magically rise over a period of time, even if the chips themselves were no different (though you needed a BIOS flash to "magically" get the higher speed). What about AMD's earlier attempts with the K5 chip (itself a footnote in Tom's article)? Speaking of the K5 chip, what about the fact that they were "like" the Athlon chips in that originally they were sold/marketed at their actual MHz, and after some optimizations they were sold/marketed using a "PR" speed? What about AMD's original? I'm speaking of the chip that was in reality a 486 core (perhaps with optimized/increased cache) running at 133MHz. Some people called it a 486DX4-133. More "officially" it was a 5x86-133 P75. Yup, AMD was calling a 486 chip on steroids a Pentium 75.
Where are all the others? The IBM/Cyrix chips? The IDT/Winchip? The NexGen 586 chip? These should have their place in history.
Comments?