mkarcher wrote on 2020-08-21, 09:08:

The "real" P75 also likely was a great competitor to the Am5x86-P75 (133 MHz), and helped getting those sales to Intel instead of AMD.

I think it was the other way around. The 5x86 took sales away from Intel. For the record, the 5x86 came out more than a year after the P75. At the time of release, the P75 had no competitors other than Intel's own P60/P66. The P75 and Intel DX4 were also both released in roughly the same timeframe (6 months apart).

Anonymous Coward wrote on 2020-08-21, 12:08:

mkarcher wrote on 2020-08-21, 09:08:

The "real" P75 also likely was a great competitor to the Am5x86-P75 (133 MHz), and helped getting those sales to Intel instead of AMD.

I think it was the other way around. The 5x86 took sales away from Intel. For the record, the 5x86 came out more than a year after the P75. At the time of release, the P75 had no competitors other than Intel's own P60/P66. The P75 and Intel DX4 were also both released in roughly the same timeframe (6 months apart).

Yep, the 5x86 was the last successful and final strike of the 486 core.

CPU clock isn't the same thing as FSB. Your CPU clock multiplier, and the bus speed are separate things. 83 mhz bus was mostly a Cyrex thing. Cyrex died thats why it wasn't a thing. Intel skipped it with the p III and clock multiplied 100 by 1x increments with a 1/3 clock devider so pci always ran at 33mhz thats why. Early boards had 1/2 dividers thats where double clocked comes from 25 or 33. VLB bandwidth was based on cpu bus speed so overclocking the VLB wasn't really a issue. IBM blue lighting is example of triple clocked 25x3.. Problems are not until PCI becomes a thing. Then some socket 7 had 1/3 dividers to support Cyrex , which resulted in socket 7 boards with cyrex chips the pci bus run at 27.7 MHz. Cyrex MII is the only chip that used 83mhz bus and if your board didn't have 1/3 divider it was unstable.

Now having proper clock dividers on boards is a common thing.

Warlord wrote on 2020-08-21, 13:06:

Cyrex MII is the only chip that used 83mhz bus and if your board didn't have 1/3 divider it was unstable.

Now having proper clock dividers on boards is a common thing.

I ran a 75mhz or 83mhz stock FSB from 1997 to 2001 without the proper divider and everything was rock solid.

Most every board you could buy to build a pc had at least a 75mhz FSB and as long as your pci was just running the video card I never had a problem.

These higher clocks became important for folks who wanted to upgrade to a k6-2 in later years after s7 was Doa.

I would honestly say I don’t know of a single 3rd party board made in the s7 era that didn’t have at least 75mhz. Oddly enough many 1997+ “value systems “ from 1st parties usually had 75mhz as well.

rmay635703 wrote on 2020-08-21, 14:17:

Warlord wrote on 2020-08-21, 13:06:

Cyrex MII is the only chip that used 83mhz bus and if your board didn't have 1/3 divider it was unstable.

Now having proper clock dividers on boards is a common thing.

I would honestly say I don’t know of a single 3rd party board made in the s7 era that didn’t have at least 75mhz. Oddly enough many 1997+ “value systems “ from 1st parties usually had 75mhz as well.

All "Super Socket 7 boards" have 1/3 dividers and its not a issue
But running a Cyrix MII with a 83FSB in a S7 board with only 1/2 divider can cause stability issues.
These are boards that are mostly for running intel MMX or intel S7s at 66mhz fsb
There are plenty of Video cards that can't handle that back in 1997, rest of the stuff like network cards or anything else u have in there can absolutely crash running at 41.5 mhz.

75mhz isnt as bad. Thats considered a slight overclock on a Pentium 233 will run it at 250mhz. That will just run the PCI bus at 37mhz which is borderline.