Hello Mad_Wolf,

Yes, I will send you the patched MD-4DUV(C) VER 3.1 BIOS via PM for you to test.

Cheers, Jan

Chkcpu wrote on 2026-01-07, 18:26:

Hello Mad_Wolf,

Yes, I will send you the patched MD-4DUV(C) VER 3.1 BIOS via PM for you to test.

Cheers, Jan

Thanks a lot m8
Is it also possible to patch both of them (Aquarius Systems/BCOM MD-4DUV - Aquarius Systems/BCOM MD-4DUVC)

Want to try both because dont know if the rom that I'm using is the correct one.

https://theretroweb.com/motherboard/bios/nm27 … 8f814185677.bin
https://theretroweb.com/motherboard/bios/md-4 … 64609169619.zip

aleodan wrote on 2025-11-14, 21:54:

Hello. I am in the process of porting award bios 6.0 instead ami bios on a PCCHIPS mainboard with SiS630 chipset. I tested 20 bioses with XGEcu, then found a very good one: all on-chip resources are working but the PCI slots are not. I had already modified the $PIR table and I think is OK, but there is no method to re-insert the decompressed and edited bios back to the compressed one, because AWDBEDIT does not recognize the structure. MODBIN cannot work, CBROM can't replace Sytem bios section either, so what can I do just to modify that small table?!

The old bios does not fully initialize the chipset (PCI transfer 12M/s, MemRead 160M/s) and the new BIOS is already great (PCI 62M/s, MEM 504M/s), but I need PCIs for a better video card and USB2.0. It is a shame to quit just because of re-insertion. Can you help me, please?!

you can replace the system bios with my tool: https://github.com/adegn73/Awd

The Awdbedit checksum issue is fixed in my tool. Awdbedit has a hardcoded checksum - my tool computes it

adegn wrote on 2026-01-26, 10:41:

aleodan wrote on 2025-11-14, 21:54:

Hello. I am in the process of porting award bios 6.0 instead ami bios on a PCCHIPS mainboard with SiS630 chipset. I tested 20 bioses with XGEcu, then found a very good one: all on-chip resources are working but the PCI slots are not. I had already modified the $PIR table and I think is OK, but there is no method to re-insert the decompressed and edited bios back to the compressed one, because AWDBEDIT does not recognize the structure. MODBIN cannot work, CBROM can't replace Sytem bios section either, so what can I do just to modify that small table?!

The old bios does not fully initialize the chipset (PCI transfer 12M/s, MemRead 160M/s) and the new BIOS is already great (PCI 62M/s, MEM 504M/s), but I need PCIs for a better video card and USB2.0. It is a shame to quit just because of re-insertion. Can you help me, please?!

you can replace the system bios with my tool: https://github.com/adegn73/Awd

The Awdbedit checksum issue is fixed in my tool. Awdbedit has a hardcoded checksum - my tool computes it

Well thank you, I am sure it is a great tool ! But the main problem was not the routing table. The two BIOSes had different address range initializations so it really was a useless attempt. Since then I already ditched that board, then repaired an old Zida ZX98-CU very damaged board from Ukraine with the Intel ZX chipset that has the memory speed from another world (780MB/s at 100MHz). All programs work fanthastic now, even some h264videos. Amasing how much of a difference could make the Intel BX/ZX chipset those days. It has some RAS precharge that masks all RAM burst accesses, is just from another world ! And in the end hardware repair was much-much easier than BIOS modding, even if I know assembly language well. Pentium III level BIOS is just too complicated !

Chkcpu wrote on 2025-11-26, 17:46:

Nice to hear that the K6-233 works fine now via that voltage adapter, and thanks for uploading the original BIOS8.ROM.

It took me a while, but I’ve just started disassembling this BIOS to see what needs to be changed for K6-2(+)/III(+) support. This is the oldest AMI BIOS I’ve ever patched, so I don’t know yet if I’m able to add this CPU support.
I’ll let you know what I find.

Hi, Jan. First, I want to say that I really appreciate you looking into this and I absolutely want to respect your time. I don't mean my inquiry with any pressure at all. However, curiosity has gotten the better of me and I can't help asking... Do you have any updates on the possibility of adding K6-2(+)/III(+) support to the PC Compatibility Card BIOS?

evanboonie wrote on 2026-02-10, 02:14:

Chkcpu wrote on 2025-11-26, 17:46:

Nice to hear that the K6-233 works fine now via that voltage adapter, and thanks for uploading the original BIOS8.ROM.

It took me a while, but I’ve just started disassembling this BIOS to see what needs to be changed for K6-2(+)/III(+) support. This is the oldest AMI BIOS I’ve ever patched, so I don’t know yet if I’m able to add this CPU support.
I’ll let you know what I find.

Hi, Jan. First, I want to say that I really appreciate you looking into this and I absolutely want to respect your time. I don't mean my inquiry with any pressure at all. However, curiosity has gotten the better of me and I can't help asking... Do you have any updates on the possibility of adding K6-2(+)/III(+) support to the PC Compatibility Card BIOS?

Hi evanboonie,

Sorry for the long silence, I have been quite busy and haven’t had much time for the BIOS patching hobby. So progress is slow, but among other projects I’m still working on your Apple PC Compatible card BIOS.

Analyzing this BIOS taught me new insights into how AMI did the CPU and Speed detection in those days. As I don’t have any ready-made patches for this older code, I have to come up with new ideas about how to fix the limitations in this BIOS.
I nice new challenge! 😉

I hope to have more soon.
Cheers, Jan

Hi Jan,

I have a couple of BIOS patch requests - the first one is for ECS P5ST-BR:
https://theretroweb.com/motherboards/s/ecs-p5st-br

I tried to use your ASUS SP97-XV BIOS because I figured the Winbond I/O chip was a minor revision difference between the two boards, but it completely freezes at the memory test in POST.

The second request is for NewTech/SMT VX3-2b:
https://theretroweb.com/motherboards/s/newtech-smt-vx3-2b

I tried to use rom.by patcher for both, but the best I could do was get 128GB HDD support on the ECS board.

DEAT wrote on 2026-02-12, 04:18:

Hi Jan, […]
Show full quote

Hi Jan,

I have a couple of BIOS patch requests - the first one is for ECS P5ST-BR:
https://theretroweb.com/motherboards/s/ecs-p5st-br

I tried to use your ASUS SP97-XV BIOS because I figured the Winbond I/O chip was a minor revision difference between the two boards, but it completely freezes at the memory test in POST.

The second request is for NewTech/SMT VX3-2b:
https://theretroweb.com/motherboards/s/newtech-smt-vx3-2b

I tried to use rom.by patcher for both, but the best I could do was get 128GB HDD support on the ECS board.

Hi DEAT,

You are in luck. I found an aftermarket BIOS upgrade by Unicore Software for both boards.
These 03/2000 upgrade Award BIOSes have all the usual goodies like K6-2(+)-III(+) and 128GiB HDD support.

Here are the BIOSes:

Using Modbin and a Hexeditor, I've checked both BIOSes against their latest original version from TRW, and they are a perfect match.
Hopefully, these Unicore BIOSes work as expected. If so, they will save me a lot off time patching the original versions. 😉

Happy testing,
Jan

Chkcpu wrote on 2026-02-12, 15:39:

Hi DEAT, […]
Show full quote

Hi DEAT,

You are in luck. I found an aftermarket BIOS upgrade by Unicore Software for both boards.
These 03/2000 upgrade Award BIOSes have all the usual goodies like K6-2(+)-III(+) and 128GiB HDD support.

Here are the BIOSes:

Thanks for the quick response! The SMT VX3-2b BIOS worked perfectly, but no dice with the ECS P5ST-BR BIOS - the same thing happens as when I've used the ASUS SP97-XV BIOS where it gets stuck at trying to initialise the memory test in POST:

I cleared the CMOS and removed the battery just to be certain.

I've noticed a lot of unicore BIOSes seem to have missing or otherwise incorrect IRQ routing tables, maybe that's worth checking

maxtherabbit wrote on 2026-02-15, 15:42:

I've noticed a lot of unicore BIOSes seem to have missing or otherwise incorrect IRQ routing tables, maybe that's worth checking

Hi maxtherabbit,

Thanks for chiming in and reminding me of the incorrect IRQ routing or PCI slots table in some Unicore BIOSes.
Actually I did check this table in the ECS P5ST-BR Unicore BIOS and it is identical to that of the original BIOS. But it would probably haven’t made any difference on this peculiar socket 7 board with 8 ISA slots and ZERO PCI slots! 😉

@DEAT,
Okay, one down and one to go. 😀
Do you have a POST analyzer card to check at what code the Unicore BIOS hangs?

Cheers, Jan

I am trying to load a via vt6421 pci option rom in my award 6.00pg bios. I added the option rom - but the bios will not load it. The same option rom loads fine in an AMI Bios - so I took a look at the disassembly and noted that AMI handles this quite a lot simpler than award.

So my idea was to bypass the additional checks to handle the option rom like AMI does

1e000:E751
2e000:E751 loc_E0E751:                             ; CODE XREF: e000:E7E9↓j
3e000:E751                 mov     edi, esi        ; esi points at the beginning option rom - now edi does too
4e000:E754                 xor     eax, eax        ; eax = 0
5e000:E757                 mov     ax, [esi+18h]   ; offset to PCIR - copy into ax
6e000:E75B                 add     edi, eax        ; add pcir offset to edi
7e000:E75E                 cmp     word ptr [esi], 0AA55h ; check that the first bytes are AA55
8e000:E763                 jnz     loc_E0E7EC      ; if not - "return"
9e000:E767                 cmp     dword ptr [edi], 52494350h ; check that the first bytes after offset are PCIR
10e000:E76F                 jnz     short loc_E0E7A9
11e000:E771                 cmp     byte ptr [bp+27Dh], 0AAh ; [i][b]replace with jmp e7a2  and some nops - EB 2F 90 90 90 to avoid dev/ven check and be more similar to AMI[/b][/i]
12e000:E776                 jz      short loc_E0E7A9
13e000:E778                 cmp     [edi+4], bx
14e000:E77C                 jnz     short loc_E0E7D4
15e000:E77E                 cmp     bx, 1000h
16e000:E782                 jnz     short loc_E0E78C
17e000:E784                 cmp     word ptr [bp+258h], 100h
18e000:E78A                 jz      short loc_E0E7A2
19e000:E78C
20e000:E78C loc_E0E78C:                             ; CODE XREF: e000:E782↑j
21e000:E78C                 mov     eax, ebx
22e000:E78F                 shr     eax, 10h
23e000:E793                 cmp     word ptr [edi+6], 78h ; 'x'
24e000:E798                 jnz     short loc_E0E79C
25e000:E79A                 xor     ah, ah
26e000:E79C
27e000:E79C loc_E0E79C:                             ; CODE XREF: e000:E798↑j
28e000:E79C                 cmp     [edi+6], ax
29e000:E7A0                 jnz     short loc_E0E7D4
30e000:E7A2
31e000:E7A2 loc_E0E7A2:                             ; CODE XREF: e000:E78A↑j
32e000:E7A2                 cmp     byte ptr [edi+14h], 0
33e000:E7A7                 jnz     short loc_E0E7D4
34e000:E7A9
35e000:E7A9 loc_E0E7A9:                             ; CODE XREF: e000:E76F↑j
36e000:E7A9                                         ; e000:E776↑j
37e000:E7A9                 push    cx
38e000:E7AA                 xor     ecx, ecx
39e000:E7AD                 mov     cl, [esi+2]
40e000:E7B1                 shl     ecx, 7
41e000:E7B5                 mov     edi, 70000h
42e000:E7BB
43e000:E7BB loc_E0E7BB:                             ; CODE XREF: e000:E7CB↓j
44e000:E7BB                 mov     eax, [esi]
45e000:E7BF                 mov     [edi], eax
46e000:E7C3                 add     edi, 4
47e000:E7C7                 add     esi, 4
48e000:E7CB                 loop    loc_E0E7BB
49e000:E7CD                 pop     cx
50e000:E7CE                 call    sub_E0E7FC
51e000:E7D1                 clc
52e000:E7D2                 jmp     short loc_E0E7F2
53e000:E7D4 ; ---------------------------------------------------------------------------
54e000:E7D4
55e000:E7D4 loc_E0E7D4:                             ; CODE XREF: e000:E77C↑j
56e000:E7D4                                         ; e000:E7A0↑j ...
57e000:E7D4                 test    byte ptr [edi+15h], 80h
58e000:E7D9                 jnz     short loc_E0E7EC
59e000:E7DB                 xor     eax, eax
60e000:E7DE                 mov     ax, [edi+10h]

61e000:E7E2                 shl     eax, 9
62e000:E7E6                 add     esi, eax
63e000:E7E9                 jmp     loc_E0E751      ; esi points at the beginning option rom - now edi does too
64e000:E7EC ; ---------------------------------------------------------------------------
65e000:E7EC

But the award bios still doesn't load the option ROM. Any help would be greatly appreciated

Best regards

Hi Jan,

Can I please ask if the qdiflash tool you linked to here (QDI Titanium IIB. v2.0 and AMD K6-III+) is compatible with the QDI P5I430TX (Titanium IB+) as well as the IB and the IIIB.

It seems to be the only place I can find the QDIflash tool, so wanted to check before I brick my mobo.

Thanks for your time and all the amazing work you do!

SwedeyTodd wrote on 2026-02-16, 14:06:

Hi Jan, […]
Show full quote

Hi Jan,

Can I please ask if the qdiflash tool you linked to here (QDI Titanium IIB. v2.0 and AMD K6-III+) is compatible with the QDI P5I430TX (Titanium IB+) as well as the IB and the IIIB.

It seems to be the only place I can find the QDIflash tool, so wanted to check before I brick my mobo.

Thanks for your time and all the amazing work you do!

Hi,

Although the QDIFLASH tool I linked was indicated by QDI to work on the Titanium IB, it also worked fine on the Titanium IIIB.
As the Titanium IB+ and IIB V2.0 boards are from the same series, I’m confident that this QDIFLASH tool works on them as well.

These boards all use the same i430TX chipset and as long as the Flash tool recognizes the EEPROM chip, QDIFLASH will work fine. If the EEPROM is not recognized, the Flash tool will not proceed with the flash, so you can just try QDIFLASH with minimal risk.

Alternately, you can always use Uniflash. This tool has a wide range of supported EEPROMs and chipsets.

Greetings, Jan

Thanks for responding so quickly I really appreciate it!

Cheers.

Chkcpu wrote on 2026-02-11, 16:36:

Hi evanboonie, […]
Show full quote

evanboonie wrote on 2026-02-10, 02:14:

Chkcpu wrote on 2025-11-26, 17:46:

Nice to hear that the K6-233 works fine now via that voltage adapter, and thanks for uploading the original BIOS8.ROM.

It took me a while, but I’ve just started disassembling this BIOS to see what needs to be changed for K6-2(+)/III(+) support. This is the oldest AMI BIOS I’ve ever patched, so I don’t know yet if I’m able to add this CPU support.
I’ll let you know what I find.

Hi, Jan. First, I want to say that I really appreciate you looking into this and I absolutely want to respect your time. I don't mean my inquiry with any pressure at all. However, curiosity has gotten the better of me and I can't help asking... Do you have any updates on the possibility of adding K6-2(+)/III(+) support to the PC Compatibility Card BIOS?

Hi evanboonie,

Sorry for the long silence, I have been quite busy and haven’t had much time for the BIOS patching hobby. So progress is slow, but among other projects I’m still working on your Apple PC Compatible card BIOS.

Analyzing this BIOS taught me new insights into how AMI did the CPU and Speed detection in those days. As I don’t have any ready-made patches for this older code, I have to come up with new ideas about how to fix the limitations in this BIOS.
I nice new challenge! 😉

I hope to have more soon.
Cheers, Jan

Hi evanboonie,

I was able to invest some more time in patching the AMI BIOS for your Apple 1997 PC Compatible Card.
I’m happy to report it is finished now, almost 4 months after your initial request. 😉

I patched the original BIOS8.ROM you sent, and here is a copy of this Patch J.1 version:

I’ve added support for the 2.2V K6 model 7, K6-2, K6-III, K6-2+, and K6-III+. I’ve also expanded the K6 Write-Allocation feature to work on the K6-2(+)/III(+) models with CXT core.
I was unable to change the CPU speed display logic for these faster K6 models, so you may see some funny speed indications when going above 233MHz.

I’m very curious what you get when testing with a K6-2(+)/III(+) CPU. Hopefully there is no other hardware or software component that prevents these CPUs from running.

Happy testing!
Jan

Chkcpu wrote on 2026-02-15, 21:48:

@DEAT, Okay, one down and one to go. :) Do you have a POST analyzer card to check at what code the Unicore BIOS hangs? […]
Show full quote

@DEAT,
Okay, one down and one to go. 😀
Do you have a POST analyzer card to check at what code the Unicore BIOS hangs?

Cheers, Jan

Yep, here is a photo of where the POST card hangs:

Hi Jan,

Quick update - QDI flash didn't seem to want to work, so used Uniflash instead like you suggested.

BIOS successfully flashed on the first attempt, then installed my K6-II 400mhz and posted no probs!

Cheers