I have stumbled upon this thread while researching my new Shuttle AV18E2 (also VT82C694T chipset). By chance I have MSI 694T Pro VER 5 too and I was running PIII-s (SL5XL with 512KB L2 Cache) for some time in it with no memory related problems. I dug up this on my HDD (I no longer have this board in use):

I don't remember doing memtest specifically, but I would remember if there were any problems with memory. At least Win 98 was running fine:

If you want I can take the motherboard out of the box again and test some things for you. I have also CTCM reports if that helps. From the board related folder I also see I was using 6309v75 BIOS and I have backup of the settings stored.

Hey @elcrys!
Thanks for sharing your experience.
The P3-s I tested was the SL6BY, a latter stepping. Could that be the cause? I don't think it's likely, but who knows?
There is probably something damaged in my board. Some microscopic damage somewhere, or an IC that has had enough, or maybe I damaged something unknowingly, or maybe the chipset is damaged... I still can't tell. But interestingly, all the other CPUs work fine. Maybe as @shamino said it is related to power draw and a transistor or another component can't handle it.
Unfortunately I don't have a P3-s anymore so I can't do more tests with it. It did work under DOS (maybe because DOS memory use is limited to 64MB?) but wouldn't bootstrap Linux.
If you can, when you find time please attach a copy of your bios rom and a pictures of your board for documenting purposes and as a reference for future tests. Maybe you have different VRMs or there's some other difference.
I'll probably get a P3-s 1.4GHz in the future and continue the saga. At the moment I fulfilled this "extreme Tualatin" project with a 440BX board and a 1.2 Tualeron@1.6Ghz (133MHz bus).

OK, no problem, I always store photos of each component:

BIOS file is "A6309VMS.750", the latest version from MSI site.

BIOS settings:

This settings can be restored using CMOS.COM v1.00 (05-05-1999 Charles Dye). Syntax:

1CMOS /LOAD cmos.txt

EDIT:
One additional hint for this motherboard. I had issues with FDD controller data corruption (floppy was practically unusable) while using ISA slot. As soon as I removed soundcard (ESS 1869, tested on many other builds) from ISA slot, all problems with FDD were gone. It took me a lot of time to figure out. Maybe this could be avoided with another soundcard, or this motherboard is simply bugged. This was one of the reasons why I left this board unused. I later replaced this build with a P4 industrial motherboard with ISA slots.

Awesome! Thanks. I'm sure this will be useful in the future.

elcrys wrote on 2023-02-24, 16:21:

EDIT:
One additional hint for this motherboard. I had issues with FDD controller data corruption (floppy was practically unusable) while using ISA slot. As soon as I removed soundcard (ESS 1869, tested on many other builds) from ISA slot, all problems with FDD were gone. It took me a lot of time to figure out. Maybe this could be avoided with another soundcard, or this motherboard is simply bugged. This was one of the reasons why I left this board unused. I later replaced this build with a P4 industrial motherboard with ISA slots.

Nice tip... it seems that it's very finicky. I don't remember if I did my tests with the ISA sound card plugged in (possibly), but I do remember that using a Tualeron running Windows I had instability problems while doing some benchmarks. Something wasn't right overall. That's why I ultimately stopped using it. Probably not the most friendly board. It's for someone with a lot of patience to tweak everything until they find the perfect balance for a specific usage. Maybe it runs Coppermines fine...

I'm yet to test a different 694T-based motherboard to see if this only affects this board or if it's because of a defective chipset design.