Timings seems more likely.
mr bios wants to run more tightly. Try increasing the ram access waitstates a little with caching on.

Or, changing the cache mode, if it is set to write thru, to write back.

Vipersan wrote on Yesterday, 10:44:

I'm still gonna ask.

No memory speed setting ?

Not really buddy..unless timing adjustments are buried deep and I haven't discovered them yet..
There are 14 pages in the bios ...I'll post them up for you to look at..

....>

..last 4

Vipersan wrote on Yesterday, 16:36:

Not really buddy..unless timing adjustments are buried deep and I haven't discovered them yet..
There are 14 pages in the bios ...I'll post them up for you to look at..

PC150012 😀

side note, which I'm unsure if it's sorta bypassed by using a amd k6/2, the 430TX chipset only caches up to 64M of memory.
Anything higher will suffer some slowdown.

anyway... play around with the first three settings if you feel like it, yes, even the EDO/FPM timing.
I don't know if they might have an effect on the SDRAM you seem to be using.

weedeewee wrote on Yesterday, 18:33:

PC150012 :-) […]
Show full quote

Vipersan wrote on Yesterday, 16:36:

Not really buddy..unless timing adjustments are buried deep and I haven't discovered them yet..
There are 14 pages in the bios ...I'll post them up for you to look at..

PC150012 😀

side note, which I'm unsure if it's sorta bypassed by using a amd k6/2, the 430TX chipset only caches up to 64M of memory.
Anything higher will suffer some slowdown.

anyway... play around with the first three settings if you feel like it, yes, even the EDO/FPM timing.
I don't know if they might have an effect on the SDRAM you seem to be using.

At this point I have nothing to lose by trying...
No wait state tweaking...and unsure if the fpm edo setting would be relevant...but hey..It took me this long and a lot of waiting..just to discover disabling the cpu cache gets results.
Is it possible or even likely that this industrial setting mrbios was never written or modified to be used with K6-III.
who knows..
certainly not me
😀

Vipersan wrote on Yesterday, 19:28:

At this point I have nothing to lose by trying... No wait state tweaking...and unsure if the fpm edo setting would be relevant.. […]
Show full quote

weedeewee wrote on Yesterday, 18:33:

PC150012 :-) […]
Show full quote

Vipersan wrote on Yesterday, 16:36:

Not really buddy..unless timing adjustments are buried deep and I haven't discovered them yet..
There are 14 pages in the bios ...I'll post them up for you to look at..

PC150012 😀

side note, which I'm unsure if it's sorta bypassed by using a amd k6/2, the 430TX chipset only caches up to 64M of memory.
Anything higher will suffer some slowdown.

anyway... play around with the first three settings if you feel like it, yes, even the EDO/FPM timing.
I don't know if they might have an effect on the SDRAM you seem to be using.

At this point I have nothing to lose by trying...
No wait state tweaking...and unsure if the fpm edo setting would be relevant...but hey..It took me this long and a lot of waiting..just to discover disabling the cpu cache gets results.
Is it possible or even likely that this industrial setting mrbios was never written or modified to be used with K6-III.
who knows..
certainly not me
😀

K6-III ? mmh I only see it reporting a K6/200. no idea though.

MR Bios's tend to have just the settings that one needs, but it tends to support/drive the hardware better from the start, like the 4 floppy drive support.
This in contrast to the AMI bios's where with the use of the correct program, like the old amisetup? (i think that's what it was called), you could change so many settings, some would work, others wouldn't do a thing.

What CPU are you actually running on that board?
also, try running the speedsys benchmark. the graph bottom right should tell you something about memory/cache throughput. Compare them with cache enabled and disabled.

You have me wondering now....
I have 2 of these mobos...one is in a desktop running win98se with my rare graphics card ...a voodoo 5500 pci.
I thought they both were fitted with the same cpu...but it may well be I am wrong..
I'll find out for sure..
cheers

checkit was less than helpful ..it claims it to be an 80486
doh!

Looks like I had this cpu mixed up with another one..
It is indeed a lowly K6 @200m ...cpuZ confirmed it.
A K6-III would be overkill for this system anyhoo.
Had to pop the original bios back in in order to run win98 and cpuz..so will have a play around with those memory settings tomorrow.
😉

My CPU is an AMD K6-III+ and I don't need to disable internal cache.

Here's a few pictures of my BIOS settings. Let me know if you need anything else.

Any possibility you're mixing high and low density memory? I always try to ensure that my memory is as matched as possible. As a test, can you try removing the high density DIMM (has only 8 or 9 chips)?

Another test... Try using an add-in IDE/SATA/SCSI card and install DOS on that. Just in case the MRBIOS isn't playing nice with the onboard IDE.

-Brian

Now I remember why I removed my K6-III..
I found it and wanted to try it.
but 3 of the pins at the extreme edge where the corner is clear (location corner) are black-ish instead of gold..
In a previous life this cpu had been exposed to something corrosive.. and so far I've found no way to clean those pins.
The CPU works just fine until placed under pressure from a heat sink..
since other chips work fine I'm not blaming the socket atm...but it looks like the gold plating on those pins has gone.
I wonder if acetic acid (vinegar) would help ?

🙁

That appears to have worked...my K6-III is currently working with a block heat sink and fan.
small paintbrush > vinegar > Iso alcohol.
But it isn't a K6-III+
This afternoon I'll try it with mrbios.
There are variations involving cpu cache ...so _may_ be relevant.
????