There are two revisions of the SiS 496/497 chipset, one that supports EDO and one that does not. Both of them will work fine with FPM and parity or non-parity memory. Unbuffered.

Post a higher resolution photo of the northbridge markings if you want to know if EDO is an option.

FP is the safe option, it will definitely work. If it's just for testing, get a bog-standard 4MB SIMM with 9 4400 chips. That will support parity, but work non-parity if needed.

As for voltage - 72p SIMM in x86 is always 5V. In more exotic hardware 3.3V can be found, but in PCs you don't need to worry about that until DIMMs roll along - and there again, in normal desktop PCs all DIMMs are 3.3V; 5V DIMMs are more of a thing in some obscure server hardware (and Macs).

Thanks, guys.

Here are some closer pics of the SiS 496/497 chipset:

If this board works again, the RAM will be used with it, not only for tests. So for now I'll look for 5 V FP unbuffered parity modules, but it will be nice if this board supports EDO RAM.

P.S. I think some early Pentium boards use 3.3 V or 5 V SIMM modules (set by jumper) and I can't call them "exotic", but maybe I'm wrong. Anyway this is year ~1995 486, 5x86 mobo and maybe it supports only 5 V SIMM modules as they're "standard" for older 486 boards.

analog_programmer wrote on 2024-02-01, 23:00:

Thanks, guys. […]
Show full quote

Thanks, guys.

Here are some closer pics of the SiS 496/497 chipset:
496-497_chips.jpg

If this board works again, the RAM will be used with it, not only for tests. So for now I'll look for 5 V FP unbuffered parity modules, but it will be nice if this board supports EDO RAM.

P.S. I think some early Pentium boards use 3.3 V or 5 V SIMM modules (set by jumper) and I can't call them "exotic", but maybe I'm wrong. Anyway this is year ~1995 486, 5x86 mobo and maybe it supports only 5 V SIMM modules as they're "standard" for older 486 boards.

There are tons of revisions of the SIS496/497 not just 2. Only the very last one support EDO, they start with P.
There were a lot of postings here on Vogons, but I cannot find a complete table.
So get FPM modules and forget about EDO.

How did you make those pictures? They are really bad.

CoffeeOne wrote on 2024-02-01, 23:24:

There are tons of revisions of the SIS496/497 not just 2. Only the very last one support EDO, they start with P. There were a lo […]
Show full quote

There are tons of revisions of the SIS496/497 not just 2. Only the very last one support EDO, they start with P.
There were a lot of postings here on Vogons, but I cannot find a complete table.
So get FPM modules and forget about EDO.

How did you make those pictures? They are really bad.

Don't worry, I'm lazy (and a bit tired right now), so I just cut the chip's images from old pictures from the day when I got the board. Now they're clean as the board itself 😀

I only see "starting P" on the last row of the chip's markings "PFA0188" and "PFD0645", but I have no idea if this is what you mean for very last revision of this chipset. The date codes on chips ("9524" and "9523") point to May 1995.

P.S. Ok, found this post in the forum and it claims that "OT" + "OR" marking are for revision B4 (or B5 in a previous comment). What it means?

analog_programmer wrote on 2024-02-01, 23:37:

Don't worry, I'm lazy (and a bit tired right now), so I just cut the chip's images from old pictures from the day when I got the […]
Show full quote

CoffeeOne wrote on 2024-02-01, 23:24:

There are tons of revisions of the SIS496/497 not just 2. Only the very last one support EDO, they start with P. There were a lo […]
Show full quote

There are tons of revisions of the SIS496/497 not just 2. Only the very last one support EDO, they start with P.
There were a lot of postings here on Vogons, but I cannot find a complete table.
So get FPM modules and forget about EDO.

How did you make those pictures? They are really bad.

Don't worry, I'm lazy (and a bit tired right now), so I just cut the chip's images from old pictures from the day when I got the board. Now they're clean as the board itself 😀

I only see "starting P" on the last row of the chip's markings "PFA0188" and "PFD0645", but I have no idea if this is what you mean for very last revision of this chipset. The date codes on chips ("9524" and "9523") point to May 1995.

P.S. Ok, found this post in the forum and it claims that "OT" + "OR" marking are for revision B4 (or B5 in a previous comment). What it means?

It's B4 => FPM only PT and PR or similar supports EDO

CoffeeOne wrote on 2024-02-02, 00:46:

It's B4 => FPM only PT and PR or similar supports EDO

Thanks for this clarification.

So I have to look for 5 V unbuffered (preferable) parity Fast Page SIMMs.

Now I have to figure out how to distinguish buffered from unbuffered memory by pictures on the RAM chips. Maybe buffered SIMM modules have some "extra" chips of another type. I think that buffered RAM is intended mainly for server machines. At least I figured out for parity - non-parity RAM visual differences from another thread in the forum.

72p SIMMs are almost never buffered.

Simple way to be sure: buffers are extra chips. If the SIMM has 8 or 9 identical chips, it's not buffered. If it has multiple smaller chips as well it may be buffered.

analog_programmer wrote on 2024-02-01, 23:00:

P.S. I think some early Pentium boards use 3.3 V or 5 V SIMM modules (set by jumper) and I can't call them "exotic"

pcchips most likely, that voltage jumper was for the DIMM slot

dionb wrote on 2024-02-02, 05:27:

72p SIMMs are almost never buffered.

Simple way to be sure: buffers are extra chips. If the SIMM has 8 or 9 identical chips, it's not buffered. If it has multiple smaller chips as well it may be buffered.

Like combination of nine 424400-70 + three 421000-70 chips?

rasz_pl wrote on 2024-02-02, 05:29:

pcchips most likely, that voltage jumper was for the DIMM slot

I have one of these and always thought that this RAM-voltage set jumper is for SIMM slots as I didn't knew there are DIMMs on 5 V. Thanks for the info.