Did some more testing and I think I narrowed it down to the AUTO Config Function in the BIOS.

For whatever reason this board does not seem to like it when that function is disabled and the DRAM settings are messed with. More aggressive settings do yield faster results in benchmarks, but result in a more unstable system.

I tested various configurations including swapping different cache and RAM chips. But anything other than BIOS defaults and AUTO Config enabled resulted in instability and crashes in Doom.

Conversely with these default settings the system does appear stable. Just played through the first 8 levels of Doom with no issues.

Shponglefan wrote on 2023-12-25, 19:35:

Did some more testing and I think I narrowed it down to the AUTO Config Function in the BIOS. […]
Show full quote

Did some more testing and I think I narrowed it down to the AUTO Config Function in the BIOS.

For whatever reason this board does not seem to like it when that function is disabled and the DRAM settings are messed with. More aggressive settings do yield faster results in benchmarks, but result in a more unstable system.

I tested various configurations including swapping different cache and RAM chips. But anything other than BIOS defaults and AUTO Config enabled resulted in instability and crashes in Doom.

Conversely with these default settings the system does appear stable. Just played through the first 8 levels of Doom with no issues.

I've found that the AMIBIOS core used on SiS 461/471 boards is full of bugs when it comes to RAM/cache timings and other performance settings. Basically, the BIOS doesn't program the chipset registers with the correct values and it causes a mess which results in sub-par performance or instability. I patched the BIOS on one of my boards (MS-4132G), if you want I can take a look at yours but it may take some time as it requires some disassembly and manual patching.

TheMobRules wrote on 2023-12-25, 20:44:

I've found that the AMIBIOS core used on SiS 461/471 boards is full of bugs when it comes to RAM/cache timings and other performance settings. Basically, the BIOS doesn't program the chipset registers with the correct values and it causes a mess which results in sub-par performance or instability. I patched the BIOS on one of my boards (MS-4132G), if you want I can take a look at yours but it may take some time as it requires some disassembly and manual patching.

That's good information to know, thank you!

Since I am trying to keep the system relatively period correct, I'm going to leave it as is. Part of the period correct fun is experiencing period correct bugs. 😁

The system doesn't perform too badly overall. I compared to a couple other motherboards I have of this era and this one was the fastest of the the three. As long as it remains stable, I'm okay with that.

I migrated the 1995 Pentium 133 build into a more compact version of the same case, and swapped its Roland SCC-1 for a Roland MPU-401AT w/ a Roland SCB-55 daughtercard.

I re-used the larger case and Roland SCC-1 in my recent 486 DX4-100 OverDrive build.

Also recently acquired a 1993 SCSI CD-ROM drive which I intend to use in the 1993 486 DX2-66 build.

Did some test fits of the builds for 1997 thru 1999. Funny how they get less complex as the years go by. 😁

1997 - Pentium II 300 (Riva 128 / Matrox M3D / Canopus Pure 3D)

1998 - Pentium II 450 (Riva TNT / Voodoo 2 SLI)

1999 - Pentium III 733 (GeForce 256 DDR)

Shponglefan wrote on 2024-01-20, 02:37:

Did some test fits of the builds for 1997 thru 1999. Funny how they get less complex as the years go by. :D […]
Show full quote

Did some test fits of the builds for 1997 thru 1999. Funny how they get less complex as the years go by. 😁

1997 - Pentium II 300 (Riva 128 / Matrox M3D / Canopus Pure 3D)

Ultimate 1997 gaming build - test fit.jpg

1998 - Pentium II 450 (Riva TNT / Voodoo 2 SLI)

Ultimate 1998 gaming build - test fit.jpg

1999 - Pentium III 733 (GeForce 256 DDR)

Ultimate 1999 gaming build - test fit.jpg

Are they all the same sound cards you move around, or do you really have that many? 🤣

H3nrik V! wrote on 2024-01-20, 07:26:

Are they all the same sound cards you move around, or do you really have that many? 🤣

I admit I cheated and used the same Diamond MX300 in the last two photos. But only because I was too lazy to fish another one out of my parts bins. 😁

But yes, I do have that many sound cards. Over a hundred at least, including multiple copies of a number of cards.