Hi
http://www.osdev.org/osfaq2/index.php/T ... e%20Family
I found this page an interesting read but a bit odd in a couple of places. First of all, about the k6 crashing problem. I've never experienced it, and on my K6 I've tried it on Win95 OSR2 and Win98. What I did get was the timing problem in K6-2s of 350MHz of above when moving to a 400MHz K6-2, the patch sorted that out.
Also, it sais that the K6-3 was in fact a cut down Athlon. I thought the only real difference between the K6-2 and 3 was the addition of the full speed L2-cache.
It doesn't meantion the K6-2+ and K6-3 Plus. These were mobile CPUs that were loved by overclockers. The k6-2+ has AFAIK 128K L2 cache on chip and the K6-3+ could be overclocked to 600MHz.
I was just wondering where you got some of this Info (not disputing it, just want more information about it).
Thanks
srg
Interesing thinks about K6 series in FAQ
Re:Interesing thinks about K6 series in FAQ
Believe I added that. At least, the K6-3 was equal to the Athlon, but with a reduced instruction decoding unit (2/cycle, athlon has 3/cycle). This was the cause for its relatively abominable performance but does explain why it would use the amount of energy it did.srg wrote: Also, it sais that the K6-3 was in fact a cut down Athlon. I thought the only real difference between the K6-2 and 3 was the addition of the full speed L2-cache.
I can't recall where I found that, but I believe it to be true.
On the K6's crashing, up to a certain patch in I believe win95 they crashed at some predictable times due to some timing thing that Windows abused (I heard something about the loopcc timing being too fast for it, but I don't know if I believe that).
Re:Interesing thinks about K6 series in FAQ
I was under the impression that the K6-III had very good performance, expecially compaired to the K6-2.Candy wrote:This was the cause for its relatively abominable performance but does explain why it would use the amount of energy it did.
Re:Interesing thinks about K6 series in FAQ
Compared to the K6-2 it was fast, but it consumed more energy per instruction. The Athlon then was faster yet, while not consuming much more energy.srg wrote:I was under the impression that the K6-III had very good performance, expecially compaired to the K6-2.Candy wrote:This was the cause for its relatively abominable performance but does explain why it would use the amount of energy it did.
Re:Interesing thinks about K6 series in FAQ
k6-iii cost them a lot more to produce than the k6-ii, really it was just a k6-2 with more cache. thats why they fell back to selling k6-2s instead of k6-3s after a while.
myself i had the 'b' stepping of the original k6 (overclocked from 200mhz to 215 or 225 or something... whooo)... the 'a' stepping apparently had some serious bugs that were fixed in the 'b' stepping...
myself i had the 'b' stepping of the original k6 (overclocked from 200mhz to 215 or 225 or something... whooo)... the 'a' stepping apparently had some serious bugs that were fixed in the 'b' stepping...
-- Stu --
Re:Interesing thinks about K6 series in FAQ
hmm, Well my 166mhz K6 ran perfectly at 200MHz. I got mine in november of 1997 so I expect I had the b setting as well, I'll have to check it out.df wrote: k6-iii cost them a lot more to produce than the k6-ii, really it was just a k6-2 with more cache. thats why they fell back to selling k6-2s instead of k6-3s after a while.
myself i had the 'b' stepping of the original k6 (overclocked from 200mhz to 215 or 225 or something... whooo)... the 'a' stepping apparently had some serious bugs that were fixed in the 'b' stepping...
EDIT Mine is Revision C 9737AJJW (so I'm guessing week 37 of 1997)