origin of FSB clocks....
- Thread starter AlNom
- Start date
Killer-Kris
Regular
I could easily be wrong but I think it has something to do with the actual quartz crystals used to create the clock in the first place. Beyond that I'm afraid I'm not much help.
bdotobdot2
Newcomer
I'll take a stab.... Backward compatibility
They didn't get to be multiples of 33 until CPUs started running at at least 33MHz. 
Anyway, I would venture a guess and say it's simply convenient, maybe PLLs are good at dividing/multiplying by 3, or it's because it gives a decent number of speed grades; 1/3, 2/3 and 3/3...
Backwards compatibility with what? The buses aren't even pin compatible with previous generations, so that's clearly not the reason.
Anyway, I would venture a guess and say it's simply convenient, maybe PLLs are good at dividing/multiplying by 3, or it's because it gives a decent number of speed grades; 1/3, 2/3 and 3/3...
Backwards compatibility with what? The buses aren't even pin compatible with previous generations, so that's clearly not the reason.
actually...
in 386, 486 and Pentium decade there was other fsb freqs in use too:
386:
- 16MHz (multipliers 1x and 2x (very rare))
- 25MHz (multipliers 1x)
486:
- 25 MHz (multipliers 1x, 2x, 3x (very rare) and 4x)
- 33 MHz (multipliers 1x, 2x, 3x and 4x)
in Pentium:
- 50 MHz
- 60 MHz
- 66 MHz
- 75 MHz (unofficial)
- 83 MHz (unofficial used in some OverDrive processors)
- 90 MHz (unofficial)
- 100MHz
and as far as I know, 100MHz FSB really is 100MHz. Not 99.9999 MHz. Of Course this depends on motherboard because it is all up to PLL and crystal which controls the clock signal.
in 386, 486 and Pentium decade there was other fsb freqs in use too:
386:
- 16MHz (multipliers 1x and 2x (very rare))
- 25MHz (multipliers 1x)
486:
- 25 MHz (multipliers 1x, 2x, 3x (very rare) and 4x)
- 33 MHz (multipliers 1x, 2x, 3x and 4x)
in Pentium:
- 50 MHz
- 60 MHz
- 66 MHz
- 75 MHz (unofficial)
- 83 MHz (unofficial used in some OverDrive processors)
- 90 MHz (unofficial)
- 100MHz
and as far as I know, 100MHz FSB really is 100MHz. Not 99.9999 MHz. Of Course this depends on motherboard because it is all up to PLL and crystal which controls the clock signal.
bdotobdot2
Newcomer
PC66, PC100, PC133....
PCI clock is 33 MHz...
That, and 33 is a pretty good increment for marketing speed grades that aren't too far apart, and memory clocks are just now matching the original pentiums in clock speed. It wasn't until the P4 speed ramp that Intel decided they needed a bit more of a boost between each line.
That, and 33 is a pretty good increment for marketing speed grades that aren't too far apart, and memory clocks are just now matching the original pentiums in clock speed. It wasn't until the P4 speed ramp that Intel decided they needed a bit more of a boost between each line.
PCI clock is 33 MHz...
That, and 33 is a pretty good increment for marketing speed grades that aren't too far apart, and memory clocks are just now matching the original pentiums in clock speed. It wasn't until the P4 speed ramp that Intel decided they needed a bit more of a boost between each line.
That, and 33 is a pretty good increment for marketing speed grades that aren't too far apart, and memory clocks are just now matching the original pentiums in clock speed. It wasn't until the P4 speed ramp that Intel decided they needed a bit more of a boost between each line.
