TheMobRules wrote on 2026-02-14, 02:02:

douglar wrote on 2026-02-14, 00:27:

I was using the default settings. Any idea what I might be able to tweak?

You need to short pins 16-17 on the front panel header. It seems this MR BIOS defaults turbo speed to low.

That did the trick for me.

Here's how it looked after setting 2111 cache timing and 0 wait state:

I get some great results using the sis.exe tool, but things are not stable. I can try some other Simms.

Are there any memory setting that are recommended?

I also noticed this when using a Promise IDE 20630 controller--

Benchmarks were about 6300KB/s read and write using the auto transfer rate, so it was doing something.

edit -- Looks like the rom also detects Appian ADI2-VL, OPTi 611-VL, & CMD 640B-VL

Why doesn't it work on my board.. Why! 🙁

appiah4 wrote on 2026-02-23, 05:59:

Why doesn't it work on my board.. Why! 🙁

There were SIS471 "green" and "non-green" variants or something. Basically, when you scroll through motherboards in retroweb with SIS471 chipset, they all have in bios strings either a single letter in addition to SIS471, or a bunch of letters after it. But they don't overlap.

Therefore i have concluded there were 2 variants of SIS471 chipsets.

The MR BIOS shared here shows "BIOS ID" as "SIS_41G". So it's probably for the "green" variant.

appiah4 wrote on 2026-02-23, 05:59:

Why doesn't it work on my board.. Why! 🙁

Check if your board has the same SuperIO chip as well.

douglar wrote on 2026-02-22, 18:55:

I get some great results using the sis.exe tool, but things are not stable. I can try some other Simms.

What kind of instability are you experiencing? I'll try to replicate on my side.

douglar wrote on 2026-02-22, 18:55:

Are there any memory setting that are recommended?

Set Write Delay to Early, it will give a considerable boost to the write speed.

My board is the Chicony CH-471A v3.0 https://theretroweb.com/motherboards/s/chicon … h-471a-ver.-3.0

SiS 85C471
SiS 85C407

I only get a black screen when I POST with this BIOS. When I verify the EPROM in a reader, it validates fine. I am reluctant to waste another EPROM because I don't have an eraser and need to put it in the sun and let it wait for 3-4 months 😁

GigAHerZ wrote on 2026-02-23, 10:56:

There were SIS471 "green" and "non-green" variants or something. Basically, when you scroll through motherboards in retroweb wit […]
Show full quote

appiah4 wrote on 2026-02-23, 05:59:

Why doesn't it work on my board.. Why! 🙁

There were SIS471 "green" and "non-green" variants or something. Basically, when you scroll through motherboards in retroweb with SIS471 chipset, they all have in bios strings either a single letter in addition to SIS471, or a bunch of letters after it. But they don't overlap.

Therefore i have concluded there were 2 variants of SIS471 chipsets.

The MR BIOS shared here shows "BIOS ID" as "SIS_41G". So it's probably for the "green" variant.

I wonder how I can check mine? I wonder if they have any markings to tell them apart. I have the Award BIOS on the board dumped, maybe looking at that could help.

EDIT: A thought just occured to me: Maybe this BIOS does not work with UMC U5S processors? That's what I have in the board currently and tested with only that. If anyone else can test their boards with a UMC CPU to confirm it works, I'd be glad.

appiah4 wrote on 2026-02-24, 07:11:

GigAHerZ wrote on 2026-02-23, 10:56:

There were SIS471 "green" and "non-green" variants or something. Basically, when you scroll through motherboards in retroweb wit […]
Show full quote

appiah4 wrote on 2026-02-23, 05:59:

Why doesn't it work on my board.. Why! 🙁

There were SIS471 "green" and "non-green" variants or something. Basically, when you scroll through motherboards in retroweb with SIS471 chipset, they all have in bios strings either a single letter in addition to SIS471, or a bunch of letters after it. But they don't overlap.

Therefore i have concluded there were 2 variants of SIS471 chipsets.

The MR BIOS shared here shows "BIOS ID" as "SIS_41G". So it's probably for the "green" variant.

I wonder how I can check mine? I wonder if they have any markings to tell them apart. I have the Award BIOS on the board dumped, maybe looking at that could help.

EDIT: A thought just occured to me: Maybe this BIOS does not work with UMC U5S processors? That's what I have in the board currently and tested with only that. If anyone else can test their boards with a UMC CPU to confirm it works, I'd be glad.

I'm afraid the "A" vs "G" is the defining indicator.

I have this board: https://theretroweb.com/motherboards/s/qdi-v4s471-g
Here's also another board with updated bios for L2 cache fix: https://theretroweb.com/motherboards/s/asus-vl-i-486sv2gx4
And i remember that on my board, the "A" BIOSes either didn't boot at all or hanged during very first moments of POST. All "G" board bioses worked well. (As a result, the bios with "write-back L2 cache" on Asus and the "EFA bios that works" on QDI are bioses submitted by me)

On the QDI's board bios list, the EFA bios that i found, notice the BIOS string: "85C471B/E/G" - no "A", but has everything else. From this i conclude there are basically 2 variants (in practice) of SIS471.

The BIOS string from mine is 01/18/95-SIS-85C471B/E/G-2C4I9C21-00 So I don't know what to make of that. The board just does not POST with the MR BIOS.

Here is the BIOS from my CH-471A V3.0 in case anyone wants to submit to theretroweb.com I really don't want the hassle of going through their Discord 😁

Okay, it now got a lot more confusing and my theory does not hold water any more...

appiah4 wrote on 2026-02-24, 18:27:

The BIOS string from mine is 01/18/95-SIS-85C471B/E/G-2C4I9C21-00 So I don't know what to make of that. The board just does not POST with the MR BIOS.

Here is the BIOS from my CH-471A V3.0 in case anyone wants to submit to theretroweb.com I really don't want the hassle of going through their Discord 😁

submitted on their discord hassle 😉
https://discord.com/channels/7131124000599572 … 953445130338632

edit: also, for some reason on https://theretroweb.com/motherboards/s/chicon … h-471a-ver.-3.0
it indicates 1 file for bios but none is available.

appiah4 wrote on 2026-02-24, 18:27:

The BIOS string from mine is 01/18/95-SIS-85C471B/E/G-2C4I9C21-00 So I don't know what to make of that. The board just does not POST with the MR BIOS.

Here is the BIOS from my CH-471A V3.0 in case anyone wants to submit to theretroweb.com I really don't want the hassle of going through their Discord 😁

Discord is preferred but you can send an email. Their email address is at the bottom of the page.

By the way, are you using a POST card?

weedeewee wrote on 2026-02-24, 20:32:

it indicates 1 file for bios but none is available.

That's strange right? A stub for a bios entry was created but never populated with any info or file. It's fixed now that your BIOS info was added to the stub.

You should try a post card with your board. MR Bios post codes are on page 102.

https://drivers.dosreloaded.de/Books/The%20BI … OST%20Codes.pdf

douglar wrote on 2026-02-25, 02:11:

That's strange right? A stub for a bios entry was created but never populated with any info or file. It's fixed now that your […]
Show full quote

weedeewee wrote on 2026-02-24, 20:32:

it indicates 1 file for bios but none is available.

That's strange right? A stub for a bios entry was created but never populated with any info or file. It's fixed now that your BIOS info was added to the stub.

You should try a post card with your board. MR Bios post codes are on page 102.

https://drivers.dosreloaded.de/Books/The%20BI … OST%20Codes.pdf

It's appiah4's bios, and whomever added it didn't checkmark the discord entry.

edit: oh, so you added it 😀
checked the logs.
I did notice that there was an entry on the v1.0 board revision that had the same bios string.

megatog615 wrote on 2026-02-24, 22:45:

Discord is preferred but you can send an email. Their email address is at the bottom of the page.

By the way, are you using a POST card?

I am using a POST card, I get no POST codes. I don't recall if the IRDA light goes on though, but IIRC it does not 🙁

Things I haven't tried:
-Try with an Intel CPU
-Try a second EPROM

Maybe this weekend I'll get to try those.

douglar wrote on 2026-02-22, 18:55:

That did the trick for me. […]
Show full quote

That did the trick for me.

The attachment TURBO_result.png is no longer available

Here's how it looked after setting 2111 cache timing and 0 wait state:

The attachment 2111-1_result.png is no longer available

I get some great results using the sis.exe tool, but things are not stable. I can try some other Simms.

Are there any memory setting that are recommended?

The attachment MemorySettings.jpg is no longer available

I also noticed this when using a Promise IDE 20630 controller--

The attachment Promise 20630.png is no longer available

Benchmarks were about 6300KB/s read and write using the auto transfer rate, so it was doing something.

edit -- Looks like the rom also detects Appian ADI2-VL, OPTi 611-VL, & CMD 640B-VL

That's neat. It never offered any IDE settings on my QD6580 card. I always wondered what it was talking about for speed settings.

Hi, so I have a Gigabyte GA-486VF, which is based on SiS 85C471/407. https://theretroweb.com/motherboards/s/gigabyte-ga-486vf

I tried the MR BIOS for it (mrbios-sis471-v3-21.zip) but it's not working properly. I can enter the BIOS and configure all of the options, but my Buslogic BT-445 throws a "HOST ADAPTER CMD PROTOCOL FAILED". This was confusing for a while and changing card settings and addresses had no impact, nor did toggling a ton of options in the BIOS itself (cache off, shadowing off, different VLB settings, etc). So I moved to an Adaptec AHA-2842A, which did get as far as the DOS boot menu but hangs as soon as you select an option.

After reviewing the MR BIOS documentation, it looks like there were multiple versions for different boards, and one main differentiator was the number of RAM slots. The BIOS I used from the forum here appears to be from a board with 4 RAM slots, while my board has 8 slots. The MR BIOS inventory seems to indicate that part number V089B415 OR V089B412 would be the correct one to use when you have 8 memory banks. And of course those have also been lost to time until someone finds it.

So I believe I'm out of luck for now. Unless I'm not understanding their inventory correctly (see attached).

Anyone know if there's a MR BIOS rom building program like cbrom or amibcp to create a rom file with option roms? Or does MR BIOS not care and just let you concatenate files into one big binary rom?

rjbrown99 wrote on 2026-05-31, 00:46:

After reviewing the MR BIOS documentation, it looks like there were multiple versions for different boards, and one main differentiator was the number of RAM slots. The BIOS I used from the forum here appears to be from a board with 4 RAM slots, while my board has 8 slots. The MR BIOS inventory seems to indicate that part number V089B415 OR V089B412 would be the correct one to use when you have 8 memory banks. And of course those have also been lost to time until someone finds it.

So I believe I'm out of luck for now. Unless I'm not understanding their inventory correctly (see attached).

I assume you must have eight 30-pin RAM slots. In this case, you have two memory banks (not eight). That would be equivalent to having two 72-pin RAM slots.