ICH5 south bridge and USB ESD protection or the lack of thereof, to be precise. Well-known issue. ASUS and several other manufacturers supposedly fixed that in new revisions.

I would refrain from using words like holocaust, Hiroshima, etc. in public forums. May hurt many people's feelings.

I have this general observation about s478 motherboards, ASUS or any other M/B brand - boards with a BGA socket 478, that in the past used a stock intel cooler with a death-clamping force, either failed or are about to fail soon due to cracking BGA joints, this fault can be simply diagnosed by turning M/B over (with a heatsink on) and pushing on the CPU area with enough force to deform PCB back a bit.

p4p800 has a bios corruption issue and also the 2 mosfets on 1.5v rail near the north bridge chipset get very hot and eventually die.

SSTV2 wrote on 2023-06-10, 17:10:

I have this general observation about s478 motherboards, ASUS or any other M/B brand - boards with a BGA socket 478, that in the past used a stock intel cooler with a death-clamping force, either failed or are about to fail soon due to cracking BGA joints, this fault can be simply diagnosed by turning M/B over (with a heatsink on) and pushing on the CPU area with enough force to deform PCB back a bit.

thats interesting. and they even had a backplate on the motherboard
cooler alternatives are also bad , since they need the stock plastic retention, like the zalman cooolers

Zerthimon wrote on 2023-06-10, 15:50:

I would refrain from using words like holocaust, Hiroshima, etc. in public forums. May hurt many people's feelings.

Agreed and thus I changed the topic

SSTV2 wrote on 2023-06-10, 17:10:

I have this general observation about s478 motherboards, ASUS or any other M/B brand - boards with a BGA socket 478, that in the past used a stock intel cooler with a death-clamping force, either failed or are about to fail soon due to cracking BGA joints, this fault can be simply diagnosed by turning M/B over (with a heatsink on) and pushing on the CPU area with enough force to deform PCB back a bit.

Socket 478 is Pin Grid Array. Typical PCB deformation from cooler won't affect it.

Here is P4P800 socket soldering. Does not look like BGA to me.

A remnant of the solder ball in the left corner indicates that it's a BGA socket. You can instantly tell which type of mounting the socket is using by turning M/B over.

I have yet to see such issue on a massive scale, because BGA under these pins are not physically connected to CPU and tension should be minimal anyway. ASUS also had backplate on premium S478. You certainly can damage them by brute forcing CPU out of a socket with closed latch though.

Anecdotally, the only board I have with such problems is Socket 939.

Interesting. I have at least one P4P800 and as far as I recall it's working fine. I believe I also have another ASUS 478 board which also works fine, don't remember the model. I don't do anything with P4 systems so I'm probably going to sell all that.

The Serpent Rider wrote on 2023-06-10, 15:25:

ICH5 south bridge and USB ESD protection or the lack of thereof, to be precise. Well-known issue. ASUS and several other manufacturers supposedly fixed that in new revisions.

You mean they fixed it with new motherboard revs or fixed it by using new ICH5 revs?

Which southbridges were affected by this anyway?

It's assumed that both Intel and motherboard manufacturers had collective fudge up.

Which southbridges were affected by this anyway?

You mean revision number? Not sure, but every working ICH5R board in my possession (including ASUS P4C800-E Revision 2.0) has SL742. Asrock LGA775 boards have SL7YC (ICH5).

The Serpent Rider wrote on 2023-06-12, 16:46:

It's assumed that both Intel and motherboard manufacturers had collective fudge up.

Which southbridges were affected by this anyway?

You mean revision number? Not sure, but every working ICH5R board in my possession (including ASUS P4C800-E Revision 2.0) has SL742. Asrock LGA775 boards have SL7YC (ICH5).

It's rather not a revision but a serial number.

First symbols of serial numbers of Intel ICH5 faulty chips:
32,33,34,40,41,42,43,44,45,46,47,48,49,4А,4В,4С

Karbist wrote on 2023-06-10, 18:53:

p4p800 has a bios corruption issue and also the 2 mosfets on 1.5v rail near the north bridge chipset get very hot and eventually die.

^this. I've been able to fix P4P800 boards by either replacing the eeprom (stock one was damaged and re-flashing did nothing) or replacing the VRMs. I've managed to repair 6-7 boards so far.

The Serpent Rider wrote on 2023-06-10, 15:25:

ICH5 south bridge and USB ESD protection or the lack of thereof, to be precise. Well-known issue. ASUS and several other manufacturers supposedly fixed that in new revisions.

I've never heard of this issue. Maybe it's because I don't own any early revision boards?

Roman555 wrote on 2023-06-12, 19:41:

First symbols of serial numbers of Intel ICH5 faulty chips:
32,33,34,40,41,42,43,44,45,46,47,48,49,4А,4В,4С

Well, in that case both of my boards are faulty as well: F32 and F40.

I have a p4p800 board, that's not posting and the south bridge is mad hot.

Roman555 wrote on 2023-06-12, 19:41:

It's rather not a revision but a serial number. […]
Show full quote

The Serpent Rider wrote on 2023-06-12, 16:46:

It's assumed that both Intel and motherboard manufacturers had collective fudge up.

Which southbridges were affected by this anyway?

You mean revision number? Not sure, but every working ICH5R board in my possession (including ASUS P4C800-E Revision 2.0) has SL742. Asrock LGA775 boards have SL7YC (ICH5).

It's rather not a revision but a serial number.

First symbols of serial numbers of Intel ICH5 faulty chips:
32,33,34,40,41,42,43,44,45,46,47,48,49,4А,4В,4С

Thanks! This is useful information 🙂

Tetrium wrote on 2023-06-13, 11:07:

Roman555 wrote on 2023-06-12, 19:41:

It's rather not a revision but a serial number. […]
Show full quote

The Serpent Rider wrote on 2023-06-12, 16:46:

It's assumed that both Intel and motherboard manufacturers had collective fudge up.

You mean revision number? Not sure, but every working ICH5R board in my possession (including ASUS P4C800-E Revision 2.0) has SL742. Asrock LGA775 boards have SL7YC (ICH5).

It's rather not a revision but a serial number.

First symbols of serial numbers of Intel ICH5 faulty chips:
32,33,34,40,41,42,43,44,45,46,47,48,49,4А,4В,4С

Thanks! This is useful information 🙂

Welcome! I got this information on a ixbt forum. But I haven't checked it personally.