Additional photos of further differences.

janskjaer wrote on 2020-07-02, 13:38:

Notice the difference in capacitor layout. For example, CE33 and CT2 are missing on the v1.03. Perhaps this has something to do with the additional ISA slot on the v1.03 or the additional PCI slot on the v1.04 at the other end of the board?

My suspicion is that the differences in the VRM layout are related to the update to VRM 8.4 that occured during the v1.03 .
The ISA slot does not seem to have anything to do with it - I have a PCI version of the v1.03 here and it has the same VRM layout as the v1.03 in your photos.

The third brand on the 1.03 is Nippon Chemicon, probably KZE series.

PARKE wrote on 2020-07-02, 14:50:

My suspicion is that the differences in the VRM layout are related to the update to VRM 8.4 that occured during the v1.03 .

Yes, only revision 1.04 of the P3B-F officially supports processors with higher power demands (i.e. most Coppermine cores), see here.

The capacitor in location CE33 is not found on all boards. There are rev. 1.04 boards with and without it. Supposedly, it was a added as a fix for certain types of instabilities; there are even reworking guides to that purpose.

DenizOezmen wrote on 2020-07-02, 18:52:

PARKE wrote on 2020-07-02, 14:50:

My suspicion is that the differences in the VRM layout are related to the update to VRM 8.4 that occured during the v1.03 .

Yes, only revision 1.04 of the P3B-F officially supports processors with higher power demands (i.e. most Coppermine cores), see here.

That is the 'official' interpretation but there must be a truckload of P3B-F boards rev 1.03 around in the wild with the 'old' VRM layout but that are accomodated with VRM 8.4 compliant voltage regulator chips. Both my PB3-F revision 1.03 boards for examplw are like that and support all Coppermines up to 1.1 Ghz.
And the rev 1.03 board in the photo posted by janskjaer has a Unisem 3007CW chip which is also VRM 8.4 compliant and that board will most likely also support all Coppermines.

PARKE wrote on 2020-07-02, 20:00:

That is the 'official' interpretation but there must be a truckload of P3B-F boards rev 1.03 around in the wild with the 'old' VRM layout but that are accomodated with VRM 8.4 compliant voltage regulator chips. Both my PB3-F revision 1.03 boards for examplw are like that and support all Coppermines up to 1.1 Ghz.
And the rev 1.03 board in the photo posted by janskjaer has a Unisem 3007CW chip which is also VRM 8.4 compliant and that board will most likely also support all Coppermines.

That's true. Many people have been running Coppermines on rev 1.03 boards, and the boards are probably built robustly enough to continue doing so.

(What is usually not officially documented by mainboard manufacturers is the maximum current draw their VRM implementation supports. They are probably designed with quite a bit of headroom, and I would guess the rev 1.04 implementation was improved in that regard as Intel released new processor steppings.)

DenizOezmen wrote on 2020-07-02, 18:52:

PARKE wrote on 2020-07-02, 14:50:

My suspicion is that the differences in the VRM layout are related to the update to VRM 8.4 that occured during the v1.03 .

Yes, only revision 1.04 of the P3B-F officially supports processors with higher power demands (i.e. most Coppermine cores), see here.

The capacitor in location CE33 is not found on all boards. There are rev. 1.04 boards with and without it. Supposedly, it was a added as a fix for certain types of instabilities; there are even reworking guides to that purpose.

Does that mean the instabilities were only evident on the v1.04 and not the v1.03?
As @PARKE has had first-hand experience at running higher specced Coppermines on a v1.03 with VRM 8.4, wouldn't that make them more reliable than the v1.04? Is that possible? For instance, what is the VRM version layout on the v1.04? Could it be in any way inferior to that of the v1.03?

janskjaer wrote on 2020-07-04, 15:38:

Does that mean the instabilities were only evident on the v1.04 and not the v1.03?
As @PARKE has had first-hand experience at running higher specced Coppermines on a v1.03 with VRM 8.4, wouldn't that make them more reliable than the v1.04? Is that possible? For instance, what is the VRM version layout on the v1.04? Could it be in any way inferior to that of the v1.03?

That does not seem likely.
The first Coppermines (with voltages lower than 1.8v) were released october 1999.
The P3B-F rev. 1.01 dates from july 1999 and the P3B-F has seen several revisions since then.
So the earlier revisions of the P3B-F were not yet optimized for Coppermines with <1.8v voltage chips and sometime during the production of revision 1.03 that transition must have taken place.
But, like Deniz wrote, it is wel possible that -some- of the revision 1.03 boards may have run into problems with Coppermines which resulted in ASUS releasing revision 1.04 (october 1999) and 1.05 (jan 2000).
Therefore the later revisions are likely more reliable than the rev. 1.03 - not less.

Another aspect is that ASUS also catered for an overclocking audience and some of the later improvements may well have focused on those users. Think for example running a 1Ghz/133 cpu with 4x256Mb of PC133 SDRAM.

janskjaer wrote on 2020-07-04, 15:38:

Does that mean the instabilities were only evident on the v1.04 and not the v1.03?
As @PARKE has had first-hand experience at running higher specced Coppermines on a v1.03 with VRM 8.4, wouldn't that make them more reliable than the v1.04? Is that possible? For instance, what is the VRM version layout on the v1.04? Could it be in any way inferior to that of the v1.03?

I didn't use a P3B-F back then, but from reading about the topic online my impression was that revision 1.03 boards were also affected. I don't know whether or not any of those were fitted with a capacitor at location CE33 from the factory, though. (Even earlier revisions of the board are said to be even more problematic, though I never saw one of those in person.)

Many people seem to have used 1.03 and 1.04 boards without noticable problems. It's likely that the stability problems only affected a subset of systems. Currently, I don't see any indication that a revision 1.04 board is inferior to a revision 1.03 board. Someone else might have different experiences, though.

PARKE wrote on 2020-07-04, 16:34:

[...] which resulted in ASUS releasing revision 1.04 (october 1999) and 1.05 (jan 2000).

Interesting that you mention this. I read about a 1.05 revision, but have never seen even a photo of it. Did you ever come across one?

When it comes down to comparing notes between individual fanciers it is imo anecdote. Even if somebody tells me that he has 2 or 3 P3B-F boards causing trouble with Coppermines it can still be the result of individual condition of the boards - or better: lack of condition.
I also have a P2B-S rev. 103 from 1998 and it runs a Coppermine Celeron 1100 just fine, go figure...

PARKE wrote on 2020-07-04, 16:48:

When it comes down to comparing notes between individual fanciers it is imo anecdote. Even if somebody tells me that he has 2 or 3 P3B-F boards causing trouble with Coppermines it can still be the result of individual condition of the boards - or better: lack of condition.

Definitely. My go-to board for PIII systems was (and is) the P3V4X, which was often derided for stability and driver problems, while I never encountered any glaring issues. It's really all down to the individual configuration.

DenizOezmen wrote on 2020-07-04, 16:44:

PARKE wrote on 2020-07-04, 16:34:

[...] which resulted in ASUS releasing revision 1.04 (october 1999) and 1.05 (jan 2000).

Interesting that you mention this. I read about a 1.05 revision, but have never seen even a photo of it. Did you ever come across one?

I have never come across one; I just know it exists because of the manual:
ftp://ftp.tekwind.co.jp/pub/asustw/mb/slot1/4 … -f/p3bf-105.pdf

PARKE wrote on 2020-07-04, 16:55:

DenizOezmen wrote on 2020-07-04, 16:44:

PARKE wrote on 2020-07-04, 16:34:

[...] which resulted in ASUS releasing revision 1.04 (october 1999) and 1.05 (jan 2000).

Interesting that you mention this. I read about a 1.05 revision, but have never seen even a photo of it. Did you ever come across one?

I have never come across one; I just know it exists because of the manual:
ftp://ftp.tekwind.co.jp/pub/asustw/mb/slot1/4 … -f/p3bf-105.pdf

That's revision 1.05 of the manual, as mentioned inside the PDF . It does not necessarily mean that a 1.05 board revision exists , IMHO .

DenizOezmen wrote on 2020-07-04, 16:55:

My go-to board for PIII systems was (and is) the P3V4X, which was often derided for stability and driver problems, while I never encountered any glaring issues. It's really all down to the individual configuration.

Oh, yes,!
P3V4X reviews triggered my curiosity in the past because I thought it was maybe the only 'pure' Coppermine board that was able to beat the P3B-F in terms of raw speed. That was not the case (here) but they are very close and the 2Mb SDRAM allows for a relatively smooth Windows XP experience 😀

darry wrote on 2020-07-04, 17:05:

PARKE wrote on 2020-07-04, 16:55:

I have never come across one; I just know it exists because of the manual:
ftp://ftp.tekwind.co.jp/pub/asustw/mb/slot1/4 … -f/p3bf-105.pdf

That's revision 1.05 of the manual, as mentioned inside the PDF . It does not necessarily mean that a 1.05 board revision exists , IMHO .

You may well be right.

PARKE wrote on 2020-07-04, 17:07:

P3V4X reviews triggered my curiosity in the past because I thought it was maybe the only 'pure' Coppermine board that was able to beat the P3B-F in terms of raw speed. That was not the case (here) but they are very close and the 2Mb SDRAM allows for a relatively smooth Windows XP experience 😀

Exactly. It really hits a sweet spot -- native 133 MHz FSB, better memory support than boards with i8xx chipsets, and it still has an ISA slot, which in later models was sacrificed for an AMR slot.

DenizOezmen wrote on 2020-07-04, 21:33:

PARKE wrote on 2020-07-04, 17:07:

P3V4X reviews triggered my curiosity in the past because I thought it was maybe the only 'pure' Coppermine board that was able to beat the P3B-F in terms of raw speed. That was not the case (here) but they are very close and the 2Mb SDRAM allows for a relatively smooth Windows XP experience 😀

Exactly. It really hits a sweet spot -- native 133 MHz FSB, better memory support than boards with i8xx chipsets, and it still has an ISA slot, which in later models was sacrificed for an AMR slot.

Except the performance - clock-for-clock - was quite a bit lower than i815 or i440BX. And that southbridge with its issues vs various PCI devices (SBLive...).

Of course Asus did make one i8xx board that has it all: the Asus P3C-E, with native 133MHz FSB, performance on par with i440BX and an ISA slot too. But rare as hen's teeth as RDRAM, which these days is no more expensive than SDRAM, but 20 years back cost far, far more. I have one, but it's dead :'(

dionb wrote on 2020-07-05, 09:03:

Of course Asus did make one i8xx board that has it all: the Asus P3C-E, with native 133MHz FSB, performance on par with i440BX and an ISA slot too. But rare as hen's teeth as RDRAM, which these days is no more expensive than SDRAM, but 20 years back cost far, far more. I have one, but it's dead :'(

But is still only goes up to 1 GB of RAM because of the i820, doesn't it?

Edit: Maybe to clarify this a bit: When I was talking about a "sweet spot", this was purely from a personal standpoint. Obviously, there is no such thing as the perfect mainboard, and everyone has to settle for their own compromise. Having worked with virtual machines early on made chipsets with a high RAM limit attractive. RDRAM was prohibitively expensive, and pairing an i820 with an MTH to use affordable SDRAM meant basically losing the performance gain over earlier Intel chipsets or those of competitors. (And in any case, the RAM limitation stayed in place.)

In the end, the i8xx chipsets never made much sense me (regarding my intended usage).