I am curious about how you make the specific pin mods. Is there a schematic you use as a reference? I'm sure it would be different for all types of motherboards.

k24a1 wrote on 2024-02-05, 18:02:

I am curious about how you make the specific pin mods. Is there a schematic you use as a reference? I'm sure it would be different for all types of motherboards.

It wouldn't, that pinmod just makes VRD11 (core 2) CPUs runs on older VRD10.1 (pentium 4/d).
VID mod is optional, but recommended! Voltage tables are totally different in VRD10.1 vs VRD11, i mean, you never can expect what voltage CPU will get, for example:
E5800 got 1.48v
X6800 got 1.56v
E2200 got 1.5v
E4700 got 1.58v
E7600 got 1.42v
All with no VID mod.

Reading through this thread has gotten me quite excited.

I have a shuttle sb77g5 which is a tiny cube based sff pc, uses lga775 socket and i875 chipset

Officially it supports p4 Prescott's only

I tried a cedarmill p4 in it and got no post at all after multiple attempts

If I've read correctly you are suggesting that if I were to inject microcode and potentially do VID pinmod I could expect functionality? Even with c2d?

Imo that combination would be awesome

Vlad94 wrote on 2024-02-01, 23:10:

You need to use afudos but engineering version 😜 I added it into attachment.
Usage: afudos /i[filename] /pbnc /n
[/quote]Nah, no success. Same as before, afudos /i1015.rom /pbnc /n just results "Romfile ROMID is not compabtible with existing bios ROMID". 🙁

Never found flashing bios this hard, do I have to modify romfile then... Asus offers beta bios 1013.001 for the Pro board though, but for the vm latest bios is 1015 with newer date. Idk if I got any chances to go with that beta one which is officially supported for my board. 😜

Bapsu wrote on 2024-02-06, 20:15:

Nah, no success. Same as before, afudos /i1015.rom /pbnc /n just results "Romfile ROMID is not compabtible with existing bios ROMID". 🙁

Never found flashing bios this hard, do I have to modify romfile then... Asus offers beta bios 1013.001 for the Pro board though, but for the vm latest bios is 1015 with newer date. Idk if I got any chances to go with that beta one which is officially supported for my board. 😜

Are you sure you replaced afudos.exe in your USB or FDD device? Maybe you didn't and tried to flash BIOS with regular version again 😜 That afudos.exe works for me even with ASRock boards.
And nope, i tried BIOS from your board earlier and it works only with Pentium 4.

tunertom wrote on 2024-02-06, 04:24:

Reading through this thread has gotten me quite excited. […]
Show full quote

Reading through this thread has gotten me quite excited.

I have a shuttle sb77g5 which is a tiny cube based sff pc, uses lga775 socket and i875 chipset

Officially it supports p4 Prescott's only

I tried a cedarmill p4 in it and got no post at all after multiple attempts

If I've read correctly you are suggesting that if I were to inject microcode and potentially do VID pinmod I could expect functionality? Even with c2d?

Imo that combination would be awesome

Wow, i didn't knew second lga775 board with i875 exists, i thought DFI 875p-t is only one. Cool!
Nope, adding microcodes wont make Core 2 initialize, it is way more complex, only way is to flash BIOS from different Core 2 board.
Obviously i875 boards for Core 2 does not exists, but i875 is almost same as i865, so it is technically possible.
Most similar board is Gigabyte GA-8I865GME-775-RH, but i can't guarantee if it will work! So flash at your own risk.
https://theretroweb.com/motherboards/s/gigaby … rh-rev-6-6#bios

Vlad94 wrote on 2024-02-07, 02:10:

Bapsu wrote on 2024-02-06, 20:15:

Nah, no success. Same as before, afudos /i1015.rom /pbnc /n just results "Romfile ROMID is not compabtible with existing bios ROMID". 🙁

Never found flashing bios this hard, do I have to modify romfile then... Asus offers beta bios 1013.001 for the Pro board though, but for the vm latest bios is 1015 with newer date. Idk if I got any chances to go with that beta one which is officially supported for my board. 😜

Are you sure you replaced afudos.exe in your USB or FDD device? Maybe you didn't and tried to flash BIOS with regular version again 😜 That afudos.exe works for me even with ASRock boards.
And nope, i tried BIOS from your board earlier and it works only with Pentium 4.

Well, I dont know what happened since I checked I had your afudos file on my usb. Erased it, downloaded whole thing again and put it back there. Did same to the biosfile itself. And it worked! 😁

Board seem to be happy with the vm version bios and everything seems to work with my current cpu. I guess next step is go ahead and find some conductive ink or paint for the e3200...

Bapsu wrote on 2024-02-08, 13:00:

Board seem to be happy with the vm version bios and everything seems to work with my current cpu. I guess next step is go ahead and find some conductive ink or paint for the e3200...

P5GD1 may not find both cores. It can happen if a signal GTLREF1 (H2 pin of CPU socket) is not routed.

P.S. Connecting H2 to H1 (GTLREF1 to GTLREF0) may help but I haven't tested it

Bapsu wrote on 2024-02-08, 13:00:

Well, I dont know what happened since I checked I had your afudos file on my usb. Erased it, downloaded whole thing again and put it back there. Did same to the biosfile itself. And it worked! 😁

Board seem to be happy with the vm version bios and everything seems to work with my current cpu. I guess next step is go ahead and find some conductive ink or paint for the e3200...

Nice to hear that BIOS worked with your board, good luck with core 2 😀

Roman555 wrote on 2024-02-08, 14:12:

P5GD1 may not find both cores. It can happen if a signal GTLREF1 (H2 pin of CPU socket) is not routed.

P.S. Connecting H2 to H1 (GTLREF1 to GTLREF0) may help but I haven't tested it

I never had such problem, c2d have cores interconnected inside Die. It always detects 2 cores for me in all boards i tried (including Asus p5gd1-vm).
But you gave me idea, i tried q6600 in Asus p5wd2-e, it detects 2 cores and freezes before BOOTing into OS, maybe connecting both GTLREF pins will fix that. I will try.

Vlad94 wrote on 2024-02-05, 22:39:

standard.png

I have a core 2 duo E4300 and IBM Thinkcentre M51 8141 with i915G chipset. I want to give a try in this mod but I'm a little bit confused.
In you picture standard.png the pinouts are mirrored of a cpu bottom. So, this mod will be like this on my CPU:

Am I right?

Vlad94 wrote on 2024-02-09, 08:06:

Roman555 wrote on 2024-02-08, 14:12:

P5GD1 may not find both cores. It can happen if a signal GTLREF1 (H2 pin of CPU socket) is not routed.

P.S. Connecting H2 to H1 (GTLREF1 to GTLREF0) may help but I haven't tested it

I never had such problem, c2d have cores interconnected inside Die. It always detects 2 cores for me in all boards i tried (including Asus p5gd1-vm).
But you gave me idea, i tried q6600 in Asus p5wd2-e, it detects 2 cores and freezes before BOOTing into OS, maybe connecting both GTLREF pins will fix that. I will try.

Maybe you are right and they all are somehow connected inside the CPU.
IMHO, contacts G10 (GTLREF3) and F2 (GTLREF2) should be routed to detect cores of the 4-core CPUs . At least connected to GTLREF0 and GTLREF1.
A chipset is also a very important part of detecting cores. Maybe you are aware that even if i945 motherboard works with C2Q it detects only 3 cores that won't run at full speed (FSB should be lower than standard otherwise the system freezes).
The last in the process is BIOS. It can spoil initialization despite a proper chipset and all signals.

tiagotarifa wrote on 2024-02-09, 11:03:

I have a core 2 duo E4300 and IBM Thinkcentre M51 8141 with i915G chipset. I want to give a try in this mod but I'm a little bit confused.

You are confused because I did STUPID mistake, thanks for noticing! I fixed it in all posts.
It should look like this:

Roman555 wrote on 2024-02-09, 11:28:

Maybe you are right and they all are somehow connected inside the CPU. IMHO, contacts G10 (GTLREF3) and F2 (GTLREF2) should be […]
Show full quote

Maybe you are right and they all are somehow connected inside the CPU.
IMHO, contacts G10 (GTLREF3) and F2 (GTLREF2) should be routed to detect cores of the 4-core CPUs . At least connected to GTLREF0 and GTLREF1.
A chipset is also a very important part of detecting cores. Maybe you are aware that even if i945 motherboard works with C2Q it detects only 3 cores that won't run at full speed (FSB should be lower than standard otherwise the system freezes).
The last in the process is BIOS. It can spoil initialization despite a proper chipset and all signals.

Are you creator of modded BIOS for this board, if yes, hello! 😁
I tried it and Q8300 crashes no matter what FSB, maybe fractional multiplier is problem...
Q6600 works fine, even OCed to 1200 FSB.
That thing with 3 cores on i945 looks like hardware bug to me, it supports Pentium D so C2Q should work fine, cuz chipset just supports SMP.

But it does not work correctly...

Vlad94 wrote on 2024-02-09, 14:57:

Are you creator of modded BIOS for this board, if yes, hello! :D I tried it and Q8300 crashes no matter what FSB, maybe fraction […]
Show full quote

Roman555 wrote on 2024-02-09, 11:28:

Maybe you are right and they all are somehow connected inside the CPU. IMHO, contacts G10 (GTLREF3) and F2 (GTLREF2) should be […]
Show full quote

Maybe you are right and they all are somehow connected inside the CPU.
IMHO, contacts G10 (GTLREF3) and F2 (GTLREF2) should be routed to detect cores of the 4-core CPUs . At least connected to GTLREF0 and GTLREF1.
A chipset is also a very important part of detecting cores. Maybe you are aware that even if i945 motherboard works with C2Q it detects only 3 cores that won't run at full speed (FSB should be lower than standard otherwise the system freezes).
The last in the process is BIOS. It can spoil initialization despite a proper chipset and all signals.

Are you creator of modded BIOS for this board, if yes, hello! 😁
I tried it and Q8300 crashes no matter what FSB, maybe fractional multiplier is problem...
Q6600 works fine, even OCed to 1200 FSB.
That thing with 3 cores on i945 looks like hardware bug to me, it supports Pentium D so C2Q should work fine, cuz chipset just supports SMP.

But it does not work correctly...

Hello! It was years ago as an experiment but the result was not bad and I got lots of fun 😀
BTW, it's an interesting idea about the hw bug that was fixed in i946.

I could try to recall that knowledge to fix 6x-multiplier in BIOS of Asus P5GD1-VM. A bootblock patch is needed so a PLC extractor, a hardware programmer (or hot-swap skill) is "must have" to resurrect a mainboard if the patch is failed

Roman555 wrote on 2024-02-10, 11:10:

Hello! It was years ago as an experiment but the result was not bad and I got lots of fun 😀
BTW, it's an interesting idea about the hw bug that was fixed in i946.

I am also surprised i946 does not have problem with quad-core, the only change is added memory remap, so it detects full 4gb, and newer integrated GPU.

Roman555 wrote on 2024-02-10, 11:10:

I could try to recall that knowledge to fix 6x-multiplier in BIOS of Asus P5GD1-VM. A bootblock patch is needed so a PLC extractor, a hardware programmer (or hot-swap skill) is "must have" to resurrect a mainboard if the patch is failed

That would be awesome! I used throttlestop to change multi to highest one, also, that BIOS does not want to load microcode for wolfdale, just gives uCode loading error.
Second issue, it does not want to POST at 800mhz fsb with Wolfdale, 667 works ok.
Third issue, E6500K, CPU with unlocked multi does not want to BOOT into OS, I guess multiplier is problem too.
P.S. It does not POST at all with dual-core Conroe for some reason, stops at code C2, single core Conroe works and it loads microcode, 800mhz FSB works too, weird.

I have USB programmer so no problem for me 😜

Vlad94 wrote on 2024-02-10, 12:16:

Second issue, it does not want to POST at 800mhz fsb with Wolfdale, 667 works ok. Third issue, E6500K, CPU with unlocked multi d […]
Show full quote

Second issue, it does not want to POST at 800mhz fsb with Wolfdale, 667 works ok.
Third issue, E6500K, CPU with unlocked multi does not want to BOOT into OS, I guess multiplier is problem too.
P.S. It does not POST at all with dual-core Conroe for some reason, stops at code C2, single core Conroe works and it loads microcode, 800mhz FSB works too, weird.

I have USB programmer so no problem for me 😜

The BIOS doesn't have Core2 specific code so it runs unpredictable way and brings strange issues trying to initialize the unknown CPU.
I doubt this issues can be fixed in reasonable period of time as they demand a lot of changes which will cause a lot of iterations (fix-flash-test). Sorry.
But I hope the x6-multi-problem can be solved pretty easy.

Roman555 wrote on 2024-02-12, 09:10:

The BIOS doesn't have Core2 specific code so it runs unpredictable way and brings strange issues trying to initialize the unknown CPU.
I doubt this issues can be fixed in reasonable period of time as they demand a lot of changes which will cause a lot of iterations (fix-flash-test). Sorry.
But I hope the x6-multi-problem can be solved pretty easy.

Well I could guess that, even oldest version from 2004 still can init Wolfdale. So obviously that BIOS just inits C2D by luck.
Even 6x multi problem fix will be ok 😀

Yiss, pinmods are my jam... checking in to catch developments. Think I skipped this era of boards where it would be useful though. There might be one Dell I could do it to, but that has numerous other incapacities that giving it more CPU wouldn't be much fun, just show them up more. (Funny about GFX, funny about RAM etc.)

One thing I would like though, would be decent CPU power with lots of regular PCI slots. So might keep an eye out for a junky board with potential for this. Think it's E2200 and E4400 I'll have spare after anything that can possibly get a quad gets one. Also got a celly 440 for a tester.

I should give the pinmod a go on one of my zd8000 boards and see if it'll work. I doubt it will initialize but it wouldn't hurt to try since I've got a unit with a faulty GPU... if it works it'll be one heck of an upgrade for it, but obviously the BIOS will be a pain in the ass for obvious reasons (Phoenix cME). There's also an unused SATA port that I'd like to get working but I'm not too sure how.

k24a1 wrote on 2024-02-12, 18:11:

I should give the pinmod a go on one of my zd8000 boards and see if it'll work. I doubt it will initialize but it wouldn't hurt to try since I've got a unit with a faulty GPU... if it works it'll be one heck of an upgrade for it, but obviously the BIOS will be a pain in the ass for obvious reasons (Phoenix cME). There's also an unused SATA port that I'd like to get working but I'm not too sure how.

Non-core 2 Award and Phoenix never worked with Core 2 for me, only AMI.
Even super old version of AMIBIOS from Asrock 775S61 works with Core 2. (I would call it socket 478 board with socket 775 soldered 🤣 ,it have gameport, but no SATA 😳)
Wolfdale freezes in BIOS after like 20 seconds, but Conroe works fine. Sadly I don't know how to add microcodes to such old AMIBIOS 🙁

Hmmm guess anything with Insyde BIOS there would be no hope as Intel would keep them on a tight leash. (Yeah you got me, that was one of those claim something on teh interwebs and wait for ppl to prove you wrong deals 🤣 )

Vlad94 wrote on 2024-02-12, 15:11:

Roman555 wrote on 2024-02-12, 09:10:

The BIOS doesn't have Core2 specific code so it runs unpredictable way and brings strange issues trying to initialize the unknown CPU.
I doubt this issues can be fixed in reasonable period of time as they demand a lot of changes which will cause a lot of iterations (fix-flash-test). Sorry.
But I hope the x6-multi-problem can be solved pretty easy.

Well I could guess that, even oldest version from 2004 still can init Wolfdale. So obviously that BIOS just inits C2D by luck.
Even 6x multi problem fix will be ok 😀

BIOS mod for Asus P5GD1-VM. I hope I didn't break it trying to fix the x6-ratio problem 😀

P.S. P5GD1-VM-mod1.zip bios mode file was removed as useless