BitWrangler wrote on 2021-07-23, 02:49:

Hi Folks, […]
Show full quote

Hi Folks,

So banging my head against this one today, though I dunno if it really matters in this instance, but it's bugging me not to know exactly why...

So I've got a DFI CA64-EN, which seems like an early FCPGA s370 board. Not a super good board I know, but you always wanna make the most of of the survivors. Anyway, it boots CC0 step PIIIs with 133 bus, a 1.65V 733 and a 1.75V 866. Latest BIOS I can see anywhere is 5/24/2001 and that's what it's got. Now if I stick a 1Ghz CD0 PIII in there, it doesn't boot, not on 133, not on 100, not on 66 (forced) while the other CPUs will force to 66 or 100 fine. So I don't think it's a power problem as it should only run at 500 forced to 66 and the 866 is working full speed on same volts. Manual and cpu-upgrade are no help, they say coppermine 533 upwards or something like that, they do not list specific chips. Now, I don't either find an update that says it supports the 100FSB PIII and gives the correct multi settings, because the settings given on board and manual top at at 8x, where it says something vague like later CPUs will override settings anyway. Well I'll tell you what happens with the CC0s if you set wrong multi, it acts stone dead, not even a fan twitch. However, that's not really the problem at the moment, as I am just inside supported multis with the 133fsb PIII, but I take it as an indication that they never updated CPU support that there's no special info for the 100fsb PIII 1Ghz, which was also a CD0 stepping. They have a number of these CA-64 boards, and when respectable manufactures might have done rev 1, 2, 3 and had BIOS updates that covered several revisions, I think they went off on the EN, EC, TC kinda kick instead.

Anyway, I can't believe that the CD0 is different enough not to boot (and give spurious model/speed etc). I do get something on the POST card when I have it set right, it either instantly goes to 88, FF or 2A at power on, turns on CPU fan. What makes me wonder though is that the CPUID code given for this CPU is 6-8-10 which is the same as for coppermine-T and early tualatins, with which the S spec intermingles... it's in the same sequence as some c-T and full T.. but this are a FCPGA-1 package, rev c, not an FCGPA-2 with a heatspreader. Now I also have a celeron coppermine, but that is also a 6-8-10 though a c in celeron revisions, but it has a 10.5 multi which I have no official info about how to set, but can extrapolate the table to guess what it might be.

So anyone know anything about CC0 and CD0 differences? Anyone know anything about CA64 revisions and whether another CA64 BIOS is basically a CA64-EN with a better BIOS BIOS?

I know right, just run it with the 733 or 866 and put the 1Ghz in one of those other boards I had running... kinda wanted to do a benchmark run with all of them at a ghz, so I got the best boards matched to the best chips.

Not sure how definitive this might be, but there's a Hot Issue page on the DFI archive and one of the topics is 'Intel Pentium III D-Stepping 1GHz/133FSB CPU prior Test List. (Jul,26,2001)'

https://web.archive.org/web/20010817122344/ht … ecentissue2.htm

The topic link (hotans16) is dead, but this appears to be the list thru another url, and the CA64-EN isn't on it !

https://web.archive.org/web/20041205150553/ht … US/HotAns16.htm

BitWrangler wrote on 2021-07-23, 02:49:

So anyone know anything about CC0 and CD0 differences?

Cannot comment on the technical aspects because that is outside of my league but the cD0 seems a different & newer/better production. Here a couple of photos:

This 850 happily runs fsb 133 at stock voltage; another one (stepping cC0) did not boot on fsb 133.

Thanks very much for those... PChoarderPatrol, I was able to dig around a bit more and found this supported CPU table...
https://web.archive.org/web/20030404165006/ht … ULstSock370.htm
Where we notice that celeron support goes to 800, PIII 100fsb support goes to 850, and 133FSB goes to 1G, but am thinking FCPGA celly and PIII/100 over those figures are the D revision and the 1G/133 officially supported is/was the C version.

However, I was able to dig around some more until I found a copy of the beta page..
https://web.archive.org/web/20021220035619/ht … te/biosbeta.asp
where the link for "Support PIII D-STEPPING CPUs." BCA64N0529.ZIP was a dead end there, but searching for that filename got me to....
https://www.elhvb.com/supportbios.info/Archiv … 4-EN/index.html
yay!
Which root domain is golden for old hardware support, but it's all dumped in sections that are hard or impossible to find from it's front page, so good thing google saw it. (Think I was there last week for a manual for this, and ended up in different archive mirrors where DFI or this board were dead end empty directories.)

Anyway, looks like that board is hooked up for getting the D 🤣 thanks for the assist.

PARKE, interesting yeah, definite difference in those components on the bottom, guess there is some electrical difference somehow. Thanks.

BitWrangler wrote on 2021-07-23, 13:14:

Thanks very much for those... PChoarderPatrol, I was able to dig around a bit more and found this supported CPU table...
https://web.archive.org/web/20030404165006/ht … ULstSock370.htm
Where we notice that celeron support goes to 800, PIII 100fsb support goes to 850, and 133FSB goes to 1G, but am thinking FCPGA celly and PIII/100 over those figures are the D revision and the 1G/133 officially supported is/was the C version.

If your observation is correct (it looks plausible) then it seems that you can identify the cpu's that work on your board via the layout of the pcb at the top of this Intel spec sheet example.
The one at the bottom represents the cD0 stepping.

PARKE wrote on 2021-07-23, 15:26:

If your observation is correct (it looks plausible) then it seems that you can identify the cpu's that work on your board via th […]
Show full quote

BitWrangler wrote on 2021-07-23, 13:14:

Thanks very much for those... PChoarderPatrol, I was able to dig around a bit more and found this supported CPU table...
https://web.archive.org/web/20030404165006/ht … ULstSock370.htm
Where we notice that celeron support goes to 800, PIII 100fsb support goes to 850, and 133FSB goes to 1G, but am thinking FCPGA celly and PIII/100 over those figures are the D revision and the 1G/133 officially supported is/was the C version.

If your observation is correct (it looks plausible) then it seems that you can identify the cpu's that work on your board via the layout of the pcb at the top of this Intel spec sheet example.
The one at the bottom represents the cD0 stepping.
P111front.JPG

I've seen a cD0 Celeron 566 SL5L5 with the older style branding.
And as others said cD0s overclock to the moon, i got that Celeron at about 1200 MHz like it was nothing.

andrea wrote on 2021-07-23, 19:47:

I've seen a cD0 Celeron 566 SL5L5 with the older style branding.

Yes, you are right. There are more Celerons like that: f.e. SL52X, SL52Y, SL54P, SL5EC, SL5LX, all from the factory in Malay.