VOGONS


Reply 20 of 25, by Anonymous Coward

User metadata
Rank l33t++
Rank
l33t++
debs3759 wrote on 2024-06-12, 15:23:

There are certainly some 90 MHz in collections, but only marked as 90 MHz, not P100 or PR100

I like to see that. The only 90's I've seen are P90+GP/80MHz.

Sphere478 wrote on 2024-06-12, 21:51:
60x1.5 90x1 45x2 […]
Show full quote

60x1.5
90x1
45x2

Some cyrix chips supported 1x but Idk if that one did.

Many around that time, only had bf0 so only two multiplier options. Which was usually 1.5 and 2x. I’m basing this on other chips, I really haven’t messed with these slower cyrix chips much. So feel free to correct. But motherboatds of that time usually had only bf0

bf1 came along around 133-166 era

Bf2 intel and bf2 amd/cyrix (different pins) came along around the 233-266 era.

Technically, it is possible to have 4 settings with one register. Pull down, pull up, floating. And a forth if you tie it to a modulating signal. Some 486 chips did this for setting I think 2 or 2.5x I think it was dx4? Tying it to breq? Signal? Trying to remember. (Don’t quote)

Anyway, having only one register suggests only 2 multiplier settings. And only 2 were typical for socket 5

6x86 and 6x86L do not support half step multipliers. Only 1X, 2X, 3X and 4X. So that rules out 1.5x60=90MHz.
If only BF0 pin is present, then probably only 2X and 3X are selectable by jumper. The other settings should be accessible by running a software utility.

"Will the highways on the internets become more few?" -Gee Dubya
V'Ger XT|Upgraded AT|Ultimate 386|Super VL/EISA 486|SMP VL/EISA Pentium

Reply 21 of 25, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

Seems only 2x and 3x are advertised to be supported on this IBM branded cyrix datasheet.

So 45mhz and 30 mhz fsb

https://datasheets.chipdb.org/IBM/x86/6x86/6x86_ALL.pdf

Seems 6x86L was a different core. Anyone have a datasheet for that?
https://en.m.wikipedia.org/wiki/Cyrix_6x86

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 22 of 25, by Anonymous Coward

User metadata
Rank l33t++
Rank
l33t++

I didn't get the impression that 6x86L uses a different core. It's just die-shrunk and split rail voltage.

"Will the highways on the internets become more few?" -Gee Dubya
V'Ger XT|Upgraded AT|Ultimate 386|Super VL/EISA 486|SMP VL/EISA Pentium

Reply 23 of 25, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
Anonymous Coward wrote on 2024-06-13, 01:46:

I didn't get the impression that 6x86L uses a different core. It's just die-shrunk and split rail voltage.

semantics. If the core shrunk and voltage split. I consider it different. Curious if cyrix kept the same core code name though?

Edit

M1 vs M1L
https://en.m.wikipedia.org/wiki/Cyrix_6 ... el%20later.

M1
M1L (Low voltage)
M1R (3M to 5M)
MII (MMX)

The wiki lists these cores, but it’s interesting as in testing we can see performance differences between the MII offerings. Those went through a number of core shrinks and apparently also some internal reconfiguring.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)