Today I tried the program on another PC, it has an i430VX chipset and a 166MHz CPU, the RAM is 32MB, unfortunately despite having tried to increase it in the past, I could not get more read.
I ran the program and created the LOG and DBG files, I am attaching them here.
The PC I used long ago to test various CPUs, at the same frequency (166), in this way I could understand what differences there were between them, I used a P.MMX a K6 and a MII, for the latter as you can see from the results I used a MII-300 lowering the frequency from 233 to 166.
If I wanted I could use other cpu, for example a K5 or an IDT-C6, but these are single voltage (it would not be a problem), the K5 works at a lower frequency around 116MHz, the IDT could perhaps have problems of recognition by the BIOS, but I remember updating it to see if it read more RAM.

PC@LIVE wrote on 2021-07-15, 21:40:

Today I tried the program on another PC, it has an i430VX chipset and a 166MHz CPU, the RAM is 32MB, unfortunately despite havin […]
Show full quote

Today I tried the program on another PC, it has an i430VX chipset and a 166MHz CPU, the RAM is 32MB, unfortunately despite having tried to increase it in the past, I could not get more read.
I ran the program and created the LOG and DBG files, I am attaching them here.
The PC I used long ago to test various CPUs, at the same frequency (166), in this way I could understand what differences there were between them, I used a P.MMX a K6 and a MII, for the latter as you can see from the results I used a MII-300 lowering the frequency from 233 to 166.
If I wanted I could use other cpu, for example a K5 or an IDT-C6, but these are single voltage (it would not be a problem), the K5 works at a lower frequency around 116MHz, the IDT could perhaps have problems of recognition by the BIOS, but I remember updating it to see if it read more RAM.

Thanks! HWiNFO has done a good job here 😀

Today I ran the program in another PC of my collection, even if the AT case is missing, I have everything necessary for operation on the bench, it is a PINE MB (if I'm not mistaken today it has become XFX), the model is PT-429, this model and I think other similar ones have a special connector for green card, in practice a special card that I don't really know what it is for, unfortunately I don't have that special card.
Currently I have 16MB of RAM, but in reality there are 20MB installed, maybe you need to change the jumper setting, but I still haven't figured out how to set them from the manual.
It has 3 memory banks, one is made up of four 30PIN modules, the other two have 72 PINs each.
The BIOS is AMI and the CPU used was initially an AMD 486-DX2-66, after running the program I replaced it with an Intel 486-DX33, I wanted to see if it was working (and indeed it was), but by the bench I have seen that the other MB with L2 cache disabled and DX2-66 gets similar results to DX-33, so if the L2 cache works it gives a nice improvement by increasing the performance.
The video card is VLB, Genoa Model 8500VL brand, has 1MB of RAM expandable to 2MB (adding 2 chips), the graphics chip is CL-GD5428.

PC@LIVE wrote on 2021-07-17, 20:30:

Today I ran the program in another PC of my collection, even if the AT case is missing, I have everything necessary for operatio […]
Show full quote

Today I ran the program in another PC of my collection, even if the AT case is missing, I have everything necessary for operation on the bench, it is a PINE MB (if I'm not mistaken today it has become XFX), the model is PT-429, this model and I think other similar ones have a special connector for green card, in practice a special card that I don't really know what it is for, unfortunately I don't have that special card.
Currently I have 16MB of RAM, but in reality there are 20MB installed, maybe you need to change the jumper setting, but I still haven't figured out how to set them from the manual.
It has 3 memory banks, one is made up of four 30PIN modules, the other two have 72 PINs each.
The BIOS is AMI and the CPU used was initially an AMD 486-DX2-66, after running the program I replaced it with an Intel 486-DX33, I wanted to see if it was working (and indeed it was), but by the bench I have seen that the other MB with L2 cache disabled and DX2-66 gets similar results to DX-33, so if the L2 cache works it gives a nice improvement by increasing the performance.
The video card is VLB, Genoa Model 8500VL brand, has 1MB of RAM expandable to 2MB (adding 2 chips), the graphics chip is CL-GD5428.

Thanks! I will add this mainboard model into my database so next time it will be recognized.
I'd also like to improve detection of the Genoa card, for that could you please run the GETBIOS tool and attach the VBIOS image (C000.ROM) ?

Mumak wrote on 2021-07-17, 20:45:

PC@LIVE wrote on 2021-07-17, 20:30:

Today I ran the program in another PC of my collection, even if the AT case is missing, I have everything necessary for operatio […]
Show full quote

Today I ran the program in another PC of my collection, even if the AT case is missing, I have everything necessary for operation on the bench, it is a PINE MB (if I'm not mistaken today it has become XFX), the model is PT-429, this model and I think other similar ones have a special connector for green card, in practice a special card that I don't really know what it is for, unfortunately I don't have that special card.
Currently I have 16MB of RAM, but in reality there are 20MB installed, maybe you need to change the jumper setting, but I still haven't figured out how to set them from the manual.
It has 3 memory banks, one is made up of four 30PIN modules, the other two have 72 PINs each.
The BIOS is AMI and the CPU used was initially an AMD 486-DX2-66, after running the program I replaced it with an Intel 486-DX33, I wanted to see if it was working (and indeed it was), but by the bench I have seen that the other MB with L2 cache disabled and DX2-66 gets similar results to DX-33, so if the L2 cache works it gives a nice improvement by increasing the performance.
The video card is VLB, Genoa Model 8500VL brand, has 1MB of RAM expandable to 2MB (adding 2 chips), the graphics chip is CL-GD5428.

Thanks! I will add this mainboard model into my database so next time it will be recognized.
I'd also like to improve detection of the Genoa card, for that could you please run the GETBIOS tool and attach the VBIOS image (C000.ROM) ?

Yes ok it's not a problem, I can do it in the next few days, on that PC I have to run other benches with the DX-33, after those I will change CPU and install an Intel Overdrive DX4-100.
For the Genoa Model 8500VL video card, I didn't quite understand what I should do, if you can write me what I should type, or what I should do, obviously after that I would have no problem creating the file you need.

Sorry, I though you already have this tool. Attaching it here:

Just run it and it will generate images of BIOS files. C000.ROM is the Video BIOS, please attach it here.

Mumak wrote on 2021-07-17, 21:13:

Sorry, I though you already have this tool. Attaching it here: GETBIOS.zip
Just run it and it will generate images of BIOS files. C000.ROM is the Video BIOS, please attach it here.

No problem, I downloaded it and I plan to run it on the PC tomorrow.
As soon as I have the file I will link it to you.
Maybe you are interested in having the files of the card with cpu overdrive?
Obviously I have to see if it works, but I'm pretty sure it works, usually the 3v ones fail if overvolted to 5v, this instead self-adjusts the voltage.

PC@LIVE wrote on 2021-07-17, 21:29:

Maybe you are interested in having the files of the card with cpu overdrive?

Yes, definitely 😀

Mumak wrote on 2021-07-17, 21:13:

Sorry, I though you already have this tool. Attaching it here: GETBIOS.zip
Just run it and it will generate images of BIOS files. C000.ROM is the Video BIOS, please attach it here.

Should that tool work for capturing all early video BIOS images? If so I'll download it, although it's still likely to be a few months before I finish reorganising my collection and clearing my work bench (going from having piles everywhere to finally putting everything on shelves and boxing all my cards individually - a seemingly never ending task at the pace I can work at, with at least a ton of parts and systems to organise)

debs3759 wrote on 2021-07-17, 21:48:

Mumak wrote on 2021-07-17, 21:13:

Sorry, I though you already have this tool. Attaching it here: GETBIOS.zip
Just run it and it will generate images of BIOS files. C000.ROM is the Video BIOS, please attach it here.

Should that tool work for capturing all early video BIOS images?

Yes, it should. Best is to run it under pure DOS (no V86).

Cool, thanks. I won't be downloading the latest HWiNFO yet, that will be updated before my test area is clear, but I have now downloaded getbios.exe to my HWiNFO directory.

Here are the files created by the getbios program, at the moment I only have these, but I can see to create others, obviously if there is something in particular that you are interested in I can create the files for you, but DOS is installed only in a few PCs, in the others I have OS from W9X and up, and I don't think it works in these

PC@LIVE wrote on 2021-07-18, 20:00:

Here are the files created by the getbios program, at the moment I only have these, but I can see to create others, obviously if there is something in particular that you are interested in I can create the files for you, but DOS is installed only in a few PCs, in the others I have OS from W9X and up, and I don't think it works in these

Thanks, that's all I needed.

Mumak wrote on 2020-04-19, 12:17:

Huh, according to the datasheet it should be F2-F3. Don't know what else to do to detect it if the result isn't as described. […]
Show full quote

Huh, according to the datasheet it should be F2-F3. Don't know what else to do to detect it if the result isn't as described.

Here a new build:
- Disabled AHA-1540 detection again.
- Automatic exclusion of SIO check for this board. You might test with 8042 and SIO check enabled
- Added FPU detection on Cyrix/TI systems without integrated FPU

[pulled, newer build available]

That AMI chip on the AMI Baby Screamer is not a VLSI 82C330 at all, but most likely just a simple memory controller with no configuration at all, other than a mysterious port 0464h that I have no idea what it does. I just did I/O port logging on my emulator using this BIOS and with shadow enabled, it does not do any I/O write that it does not do with shadow enabled. Also, no 82C330 writes, only 82C331 and what looks like a SMC Super I/O chip (begins by writing 55h twice to port 03F0h) but might be C&T 82C311 instead.

Edit: The C&T 82C311 is the CACHE/DRAM controller. Perhaps that AMI chip is the Super I/O then and is a rebadged SMC?

Edit #2: That is most likely the case, the important chips I can see, are:
C&T 82C311 - CACHE/DRAM controller;
VLSI VL82C331 - ISA BUS controller;
VLSI VL82C332 - Data buffer;
AMI Megatrends MG-9275 - Unknown;
Intel P8042AHP with AMI KB-BIOS-VER-F - Keyboard controller.

The BIOS talks to the Super I/O chip on ports 03F0 as if it was a SMC, and the only chip I can see that could be that, is that AMI Megatrends MG-9275.

Edit #3: Yes, confirmed - the 82C311 is where the Shadow RAM and other stuff is. Also, the chip markings are even swapped - more likely than not, the small chip (with the C&T 82C311) is actually the Super I/O (consistent with the facts that early SMC's were exactly of that shape), and the bigger, square chip (with the AMI marking) is the actual 82C311 (as all other photos of the 82C311 I could find, show it as big and square).

Battler wrote on 2021-07-31, 04:56:

That AMI chip on the AMI Baby Screamer is not a VLSI 82C330 at all, but most likely just a simple memory controller with no conf […]
Show full quote

Mumak wrote on 2020-04-19, 12:17:

Huh, according to the datasheet it should be F2-F3. Don't know what else to do to detect it if the result isn't as described. […]
Show full quote

Huh, according to the datasheet it should be F2-F3. Don't know what else to do to detect it if the result isn't as described.

Here a new build:
- Disabled AHA-1540 detection again.
- Automatic exclusion of SIO check for this board. You might test with 8042 and SIO check enabled
- Added FPU detection on Cyrix/TI systems without integrated FPU

[pulled, newer build available]

That AMI chip on the AMI Baby Screamer is not a VLSI 82C330 at all, but most likely just a simple memory controller with no configuration at all, other than a mysterious port 0464h that I have no idea what it does. I just did I/O port logging on my emulator using this BIOS and with shadow enabled, it does not do any I/O write that it does not do with shadow enabled. Also, no 82C330 writes, only 82C331 and what looks like a SMC Super I/O chip (begins by writing 55h twice to port 03F0h) but might be C&T 82C311 instead.

Edit: The C&T 82C311 is the CACHE/DRAM controller. Perhaps that AMI chip is the Super I/O then and is a rebadged SMC?

Edit #2: That is most likely the case, the important chips I can see, are:
C&T 82C311 - CACHE/DRAM controller;
VLSI VL82C331 - ISA BUS controller;
VLSI VL82C332 - Data buffer;
AMI Megatrends MG-9275 - Unknown;
Intel P8042AHP with AMI KB-BIOS-VER-F - Keyboard controller.

The BIOS talks to the Super I/O chip on ports 03F0 as if it was a SMC, and the only chip I can see that could be that, is that AMI Megatrends MG-9275.

Edit #3: Yes, confirmed - the 82C311 is where the Shadow RAM and other stuff is. Also, the chip markings are even swapped - more likely than not, the small chip (with the C&T 82C311) is actually the Super I/O (consistent with the facts that early SMC's were exactly of that shape), and the bigger, square chip (with the AMI marking) is the actual 82C311 (as all other photos of the 82C311 I could find, show it as big and square).

Thanks for your deep analysis 😀 It's not easy to recognize chips that don't contain any ID registers or specific functions that could be used for identification.

I stand corrected, that's not an 82C311 but an 82C711, which is an I/O chip that, for some reason, uses the same protocol as SMC. This means the MG-9275 remains a mystery - someone in this thread probed the VL82C330 ID registers on it and got FF, so it must be something else. The question is what.

I have just run the latest version on a rather unreliable Toshiba T2130CS laptop and it hangs at this point with the hard drive LED on:

SteveC wrote on 2021-08-18, 22:32:

I have just run the latest version on a rather unreliable Toshiba T2130CS laptop and it hangs at this point with the hard drive LED on:

You'll need to disable the "Check Keyboard Controller and Super IO" option there.

Isnt possible just make better code and anticipate such hard freeze.. and disable it automatically if problematic HW detected?

I can try that if I get the HWiNFO report + debug file. If there are some markers that could be used, it would be possible.