Do you need all that ram in a system like that? Have you tried to update the bios?
Also you can almost always lower the cpu multiplier. The multiplier lock prevents you from increasing the multiplier and not decreasing it.

PcBytes wrote:

I tried a 256MB double sided PC133 PQi (see my post above yours) and I just got the no RAM beeps. And the FSB was set to the old position,100MHz.

Try the misbehaving RAM modules in another board to see if they still work.

saturn wrote:

Do you need all that ram in a system like that? Have you tried to update the bios?
Also you can almost always lower the cpu multiplier. The multiplier lock prevents you from increasing the multiplier and not decreasing it.

If memory doesn't fail me, locked P2/P3 parts can neither increase nor decrease the multiplier. They are stuck there.

alexanrs wrote:

saturn wrote:

Do you need all that ram in a system like that? Have you tried to update the bios?
Also you can almost always lower the cpu multiplier. The multiplier lock prevents you from increasing the multiplier and not decreasing it.

If memory doesn't fail me, locked P2/P3 parts can neither increase nor decrease the multiplier. They are stuck there.

I never had a problem lowering them. But I always use ABIT motherboards in all my systems so who knows.
I lowered the multiplier on a pair of 1ghz P3's the other day with my vp6. I was testing to see if my cpu, ram or something else was holding back my overclock.

Weird, I googled a bit now and every reference I could find claims Coppermine P3s are fully locked (Coppermine Xeons aren't). I did see one person mentioning ABIT found out a way to unlock later P2s by setting them to 66MHz and then to 100MHz during the booting process, but that the trick didn't work on later processors.

saturn wrote:

I never had a problem lowering them. But I always use ABIT motherboards in all my systems so who knows.
I lowered the multiplier on a pair of 1ghz P3's the other day with my vp6. I was testing to see if my cpu, ram or something else was holding back my overclock.

In all my testing and reading: Retail Pentium II/III made after august 1998 are fully locked, they have only their default multiplier setting. There is nothing a motherboard can do about it.
If you are still confident, be sure to start a topic about it with screenshots/photos/Intel-product-codes.

PcBytes wrote:

I tried a 256MB double sided PC133 PQi (see my post above yours) and I just got the no RAM beeps. And the FSB was set to the old position,100MHz.

I was aware of your observation. PC133 is just a rating, the stick should also run at PC66/PC100 timings just fine. It just doesn't cooperate with your motherboard for unknown reason.

Tetrium wrote:

PcBytes wrote:

I tried a 256MB double sided PC133 PQi (see my post above yours) and I just got the no RAM beeps. And the FSB was set to the old position,100MHz.

Try the misbehaving RAM modules in another board to see if they still work.

I also did that,with a VIA 694X board,and the modules worked fine there.

gerwin wrote:

Then there are Pentium III'EB' versions that are designed for the 133MHz FSB.

The 133MHz FSB versions only got the "EB" designation when there was also a 100MHz FSB version at the same CPU speed.
Like:
The 1000MHz P-III came without EB for the 100MHz FSB version and with the EB for the 133MHz version. (As you said)
But the 667, 733, 866 and 933MHz CPUs only came in 133MHz FSB and did not get the EB because there was no 100MHz FSB version at those speeds.
Oddly the 533 got the EB even though there was no 100MHz FSB 533.
.
I think they do things like that just to confuse people.
.

PcBytes wrote:

I just tried two double-sided PC133 sticks. No go,one shows as 128 MB (a 512MB Infineon stixk out of a HP) and the other one (256MB PQi) would give the no RAM beep.

The term "double-sided" is misleading because people take it to mean there are chips on both sides of the memory stick.
What it's SUPPOSED to mean is that it's a two rank module. (Rank meaning an addressable 64-bit wide bank.)
(I use the phrasing that double vs single sided is how the modules are 'wired' and not where the chips are located on the module.)
(I prefer the term 'rank' because 'bank' has other meanings in some places.)
.
I have run across two rank (double-sided) RAM that only had chips on one side of the module.
I've also seen modules with chips on both sides that were one rank (single-sided).
I've seen such modules in both 72-pin and 168-pin RAM. Those ARE unusual and rare but they happen.

The industry has been trying to get away from using the terms "double-sided" and "single-sided" for a long while now but the terms have been in use so long they won't fade away.
.
.
Attached is the blurb on RAM for the 440BX from Intel's documentation.
[ As you can see - back in the day even Intel was using the popular misleading terms. ]
As I understand it the part where it says "8 rows memory" translates in modern terms to that it can support 8 ranks total.

This one says it can use high or low density RAM so long as the correct chips are used on the module.
It also says (assuming a 4 RAM slot board) 4x 128Mb modules (512Mb) are the max size unless you are using Registered RAM, then 4x 256Mb works.
I've played with some Intel N440BX Nightshade boards and I found that to be exactly correct on those boards.
At the time I suspected it was being enforced by the BIOS so there may be work-arounds to get non-Reg'ed 256Mb modules to work.

.

alexanrs wrote:

Weird, I googled a bit now and every reference I could find claims Coppermine P3s are fully locked (Coppermine Xeons aren't). I did see one person mentioning ABIT found out a way to unlock later P2s by setting them to 66MHz and then to 100MHz during the booting process, but that the trick didn't work on later processors.

So does this mean that Katmai P3s are locked as well? My 500MHz P3 is a Katmai.

P3 Katmai should be locked as well. Xeons and engineering samples should allow lowering multipliers.

alexanrs wrote:

Weird, I googled a bit now and every reference I could find claims Coppermine P3s are fully locked (Coppermine Xeons aren't). I did see one person mentioning ABIT found out a way to unlock later P2s by setting them to 66MHz and then to 100MHz during the booting process, but that the trick didn't work on later processors.

All I know it that my old slot one system with an abit BE62 and my vp6 system both are able to lower the cpu multi.

alexanrs wrote:

P3 Katmai should be locked as well. Xeons and engineering samples should allow lowering multipliers.

Well,my Katmai goes up to 560MHz if I change the FSB switches to 133MHz and use PC100 SDRAM. I'll try lowering the FSB and post some results later.

saturn wrote:

alexanrs wrote:

Weird, I googled a bit now and every reference I could find claims Coppermine P3s are fully locked (Coppermine Xeons aren't). I did see one person mentioning ABIT found out a way to unlock later P2s by setting them to 66MHz and then to 100MHz during the booting process, but that the trick didn't work on later processors.

All I know it that my old slot one system with an abit BE62 and my vp6 system both are able to lower the cpu multi.

Have you tested them with a bunch of different processors or just a few? Engineering samples are also unlocked (downards only), so that could be the reason your faster processors can lower the multipliers.
Anyway, if you verify that you can lower multipliers on any P3, please open a thread so everyone can investigate. I'd LOVE to have a P3 system that is able to lower its multiplier.

PcBytes wrote:

alexanrs wrote:

P3 Katmai should be locked as well. Xeons and engineering samples should allow lowering multipliers.

Well,my Katmai goes up to 560MHz if I change the FSB switches to 133MHz and use PC100 SDRAM. I'll try lowering the FSB and post some results later.

if your having stability problems try uping your ram timings. if the ram runs at CL2 set it to CL3. you should be able to get a higher FSB with the ram running at a higher CL or timmings. Also if you motherboard supports try changing your PCI and ram divider around. and don't forget to up the voltage.

alexanrs wrote:

Have you tested them with a bunch of different processors or just a few? Engineering samples are also unlocked (downards only), so that could be the reason your faster processors can lower the multipliers.
Anyway, if you verify that you can lower multipliers on any P3, please open a thread so everyone can investigate. I'd LOVE to have a P3 system that is able to lower its multiplier.

Well I don't recall every cpu I triyed. I have used engineering samples in the past and they were fully unlocked. Right now I only been using 1ghz P3's none of witch are engineering samples If I recall I did have a Gigabyte board that did not let me change the multipliers on the same P3s I used in the abit board. I'll give it another shot tonight, I know I was running the 2 1ghz P3's in my vp6 at a lower multi the other day.

alexanrs wrote:

saturn wrote:

alexanrs wrote:

Weird, I googled a bit now and every reference I could find claims Coppermine P3s are fully locked (Coppermine Xeons aren't). I did see one person mentioning ABIT found out a way to unlock later P2s by setting them to 66MHz and then to 100MHz during the booting process, but that the trick didn't work on later processors.

All I know it that my old slot one system with an abit BE62 and my vp6 system both are able to lower the cpu multi.

Have you tested them with a bunch of different processors or just a few? Engineering samples are also unlocked (downards only), so that could be the reason your faster processors can lower the multipliers.
Anyway, if you verify that you can lower multipliers on any P3, please open a thread so everyone can investigate. I'd LOVE to have a P3 system that is able to lower its multiplier.

I actually tried lowering multipliers on my 500MHz Katmai.

66 and 83MHz yield 333MHz CPU speed for 66MHz FSB and 416MHz for 83MHz. That's all I could get. Of course,133 gave me 560MHz.

PcBytes wrote:

133 gave me 560MHz.

You must have set it to 112 MHz FSB there.

gerwin wrote:

PcBytes wrote:

133 gave me 560MHz.

You must have set it to 112 MHz FSB there.

FSB*X=Clock speed. We're X is your cpu multiplier.

PcBytes wrote:

I actually tried lowering multipliers on my 500MHz Katmai.

66 and 83MHz yield 333MHz CPU speed for 66MHz FSB and 416MHz for 83MHz. That's all I could get. Of course,133 gave me 560MHz.

Changing the CPU multiplier is not the same thing as changing the FSB multiplier.
Changing the CPU multiplier means you can get different CPU speeds -without- changing the FSB.
.
That's why unlocked processors are such a big deal.
.

saturn wrote:

gerwin wrote:

PcBytes wrote:

133 gave me 560MHz.

You must have set it to 112 MHz FSB there.

FSB*X=Clock speed. We're X is your cpu multiplier.

The lowest clocked P3 450 with the bus overclocked to 133% would run at 600 Mhz. This doesn't fit.
A P3 500 with the bus overclocked to 112% would run at 560 Mhz. This does fit.
A P3 500 with the bus overclocked to 133% would run at 667 Mhz.