AlessandroB wrote on 2020-04-27, 18:50:

I have a K6 III 400 cpu […]
Show full quote

I have a K6 III 400 cpu

This one:
http://www.cpu-world.com/CPUs/K6-III/AMD-K6-I … III-400AHX.html

When i set the board to 6x multiplier the result is 266Mhz (66Mhz x 4).... why????? Tnks

The mainboard is this one:

How did you check, that the result is 266MHz?

The manual says to use 1-2 closed, 3-4 open, 5-6 open on JP2 for non-intel CPUs - is that the setting you used? If that results in 4x, have you tried the intel 2x setting (1-2 closed, 3-4 open, 5-6 closed)?

CoffeeOne wrote on 2020-04-27, 19:21:

AlessandroB wrote on 2020-04-27, 18:50:

I have a K6 III 400 cpu […]
Show full quote

I have a K6 III 400 cpu

This one:
http://www.cpu-world.com/CPUs/K6-III/AMD-K6-I … III-400AHX.html

When i set the board to 6x multiplier the result is 266Mhz (66Mhz x 4).... why????? Tnks

The mainboard is this one:

How did you check, that the result is 266MHz?

Computer post it after the boot and i see the same in speedsys

Socket3 wrote on 2020-04-27, 19:26:

The manual says to use 1-2 closed, 3-4 open, 5-6 open on JP2 for non-intel CPUs - is that the setting you used? If that results in 4x, have you tried the intel 2x setting (1-2 closed, 3-4 open, 5-6 closed)?

I have tried both and strangely both are with 266Mhz (4x)

That's weird indeed.
I was posting that, because I was running a K6-2 400 with multiplier 2 (meaning 6 for K6-2) on a Gigabyte 586-HX2. The BIOS of board reported 266MHz, too. But it was in reality the correct 400MHz, so only reported wrongly.
But here it should be different, because K6-2 and K6-III are explicitely mentioned in the manual (using multiplier 6)

What if you try out by yourself all possible combinations of JP2?
It is just 8 (3bit of information) 😁
Maybe the manual is just faulty.
You cannot damage something, because you underclock the cpu (except when you have set multi 6), so you should get all possible values from 2.5 (166MHz) up to 6 (400MHz)

I tested a K6-2 / 350afr (100x3.5 from factory) that i have.

all the multipliers are correct and work as they are written in the manual I have attached.

This CPU works very well at 400Mhz (66Mhz x6). The curious thing is that setting it to 55Mhz x 2.5 = 137Mhz you get interesting results: disabling the L1 with Speedsys you get the same speed as a 486DX33 and the timedemo of quake1 contained in DOSBENCH gives a result of 3.8fps. Interesting CPU ranging from 486DX33 to K62 400.

The problem with the K6III remains .. why don't I enable the 6x multiplier ????

Unexpected SUCCESS!

I found out that with a hot-reboot (ctrl-alt-del) the 6x multiplier is enabled ... the only thing that makes me think is that the empty CR2032 battery causes this strange behavior.

I tried again speedsys at 55x2.5 without L1 and without L2 and I got a value of 6.83 just less than a 386DX40, excellent for problematic games.

AlessandroB wrote on 2020-04-28, 13:32:

Unexpected SUCCESS!

I found out that with a hot-reboot (ctrl-alt-del) the 6x multiplier is enabled ... the only thing that makes me think is that the empty CR2032 battery causes this strange behavior.

I tried again speedsys at 55x2.5 without L1 and without L2 and I got a value of 6.83 just less than a 386DX40, excellent for problematic games.

Awesome, Have you tried it with Wing Commander 1?

Low cmos batteries can cause lots of screwy behaviour.

synrgy87 wrote on 2020-04-28, 18:44:

AlessandroB wrote on 2020-04-28, 13:32:

Unexpected SUCCESS!

I found out that with a hot-reboot (ctrl-alt-del) the 6x multiplier is enabled ... the only thing that makes me think is that the empty CR2032 battery causes this strange behavior.

I tried again speedsys at 55x2.5 without L1 and without L2 and I got a value of 6.83 just less than a 386DX40, excellent for problematic games.

Awesome, Have you tried it with Wing Commander 1?

Low cmos batteries can cause lots of screwy behaviour.

not yet... i have two little child 😀