You don't need to set any multiplier and it doesn't matter what the max multiplier is on your motherboard.
All P3s have a factory locked multiplier, even if you wanted to force another multiplier, you couldn't. Now, the cool part is that your CPU might just work at 133 x 6, which would make it an 800EB 😀. If not, chances are it will work at 124 x 6 or at least 112 x 6.

Since we are talking about a Pentium 3 Coppermine, the only problem you might have is if the motherboard doesn't support your CPU's voltage (typically 1.65V / max 1.75V) and/or you have an older BIOS version that has no Coppermine support.
Most of my 440BX motherboards support Coppermines, but I do have a few which don't.

jedikwon wrote on 2021-05-19, 06:09:

Hello, […]
Show full quote

Hello,

I have this motherboard, HOT-661V31, and wondering if I can use a Pentium III 600E CPU on this motherboard.

The manual says it only supports Pentium II and Celeron CPUs but since the chipset is 440BX, I may have a change to use a PIII CPU on it.

The problem is that the manual jumper setting states that it can do multiplier of up to 5. It seems that it does up to 150MHz FSB so may be I can set 133MHz and 4.5 multiplier for 600MHz?

But then the particular CPU I am trying to use is SL43E and uses 100MHz FSB only. Does that mean I have to overclock it?

I am sorry but I really have no idea how to use 600MHz CPU on this motherboard. Please give me any idea you have. I am attaching the manual for this motherboard. Please take a look.

Thank you.

As far as I remember revision 3.1 "should" support coppermines OOTB (your CPU is Coppermine Pentium 3). It all depends on the voltage regulating circuit (aka VRM).
Multiplier doesn't matter so much, because this CPU is multiplier locked.

Here Shuttle HOT-661 VRM upgrade gerwin upgraded VRM on older PCB revision.

bios readme:
6610S144 :

1. Fix when using ATA 66 HD can't enable DMA mode in WIN95/98
2. Update DMI support PIII CPU
3. Update DMI information "CPU MAX SPEED" from 500 to 900
4. Add "LOCK" function in item "CPU Clock Rate" and set it to default, then item
"CPU SPEED" will be disabled
5. Update CPU code (PIII 550Mhz CPU)
6. Fix when using PIII 550 CPU display speed error
7. Fix when using WD HD cannot be detected when WIN95/98 restart
8. Fix "Wake On Lan" can not work when using 3com lan card with LDCM

6610S151 :
1. fix CHIPSET FEATURE have garbage

6610S162 :
1. IBM 15GB HD from LBA mode to CHS mode after FDISK
2. 3D LAB AGP card can't display
3. add "keyboard 98" power on function
4. update CPUCODE
6610s164:
1. IBM 37.5GB support
2. Coppermine CPU support

UH19 page:
http://www.win3x.org/uh19/motherboard/show/4469

Thank you so much for the replies guys!! It looks like it will work with right BIOS. Maybe I go look for even faster CPUs like 800MHz.

SL43E is a cB0 stepping. I had good luck with those.
Many years ago I ran two SL43E CPUs in separate machines at 800/133MHz with no problem. They both passed memtest86 at 840/140 at default voltage, which was enough to convince me they were solid for 133FSB so I didn't try any higher.
What I don't remember is if I had any others that failed to take that overclock. But I think I just had those 2, so if that's correct then it was 2 for 2 success running at 800MHz.

At first I had an older cA2 stepping 600E that did not like to overclock as much. But cB0 chips will go faster and are a good bet for 800MHz.
The remaining issues then are making sure the PCI and AGP frequency dividers are set correctly, and whether your AGP card will mind 89MHz.

Something I remember about unofficial 133 support back in the day, was it was best to go with 7ns SDRAM for best stability, 7.5ns is probably rebinned 8ns, and is right near the edge at 133, so you don't want two parts near the edge, both your chipset AND the SDRAM.... However, it need not necessarily be marked PC-133, there were 7ns PC-100 modules around that ran great all the way up to the 140s.

BitWrangler wrote on 2021-05-20, 15:42:

Something I remember about unofficial 133 support back in the day, was it was best to go with 7ns SDRAM for best stability, 7.5ns is probably rebinned 8ns, and is right near the edge at 133, so you don't want two parts near the edge, both your chipset AND the SDRAM.... However, it need not necessarily be marked PC-133, there were 7ns PC-100 modules around that ran great all the way up to the 140s.

Finding PC 133 DIMMS was a bit challenging last time I tried it. Make sure not to get the ECC models confused with the non ECC ones. I have a couple of PC-100 DIMS that didn't seem to mind running at CAS 2 133mhz, but after running into some stability problems here and there and not wanting to fuss with it I lowered them down to CAS 3.

I guess it's low density PC-133 that's more the problem. Mostly just socket A boards liked the high density. Usually 128MB sticks aren't too much a problem, it's the 256 you've gotta check out well, or 512s, well you'd know a low density 512, it'd be twice the height to get all the chips on it.... and with those you get into needing to be sure you've got enough banks per slot, or that there's enough drive strength to handle them.

The BX won't be the problem, just have to be sure for BIOS and voltage support.

My P3B-f has a 4x PCI divider and P3-933 running at 980MHz with 140MHz fsb and Ti4200. This card is totally fine with the 89MHz AGP bus.
I ran a 440MX for a while at 150MHz fsb, which takes the AGP bus up to 100MHz. Not a problem at all for this card but the Ti won't go that far (but overall system speed is still better with the To). PCI was still ok, just out of spec.

In my experience the Nvidia cards are mostly tolerant to extreme AGP bus speeds, ati far less and Matrox is a no go.

And as said: cherry picked the 7ns ram module, some 7ns wouldn't even work at 150MHz.

So past 2 months, I've tried following CPUs.

PIII 800MHz SL4BX
PIII 866 MHz SL47S
PIII 600MHz SL3JP

First 2 are Coppermine CPUs and they don't even boot. =/
My motherboard BIOS does not have any function related to control voltage manually.

The last one, 600MHz, is a Katmai CPU and it runs fine on the motherboard.
BIOS updates shows it supports Coppermine CPUs so I think the problem is the voltage as someone mentioned above.

According to this link, Shuttle HOT-661 VRM upgrade , it seems that I need hip6004BCB chip instead of hip6004ACB chip which is currently installed on my motherboard for voltage issue.
Well, sadly, I have no such skill to do that.

And there I found out the existence of this Socket 370 to Slot 1 adapter. It seems promising and I can easily find over 800 MHz Socket 370 CPUs so I want to try it.
However, if the voltage is the problem there is a high chance that the adapter will not work at all.

One adapter I found shows me "MANUAL VCORE" with jumpers to set voltages from 1.60~2.40.
I don't know if this adapter can really regulates voltages or it just wants me to set correct voltage that written on CPU.

So do you guys think these Socket 370 to Slot 1 adapters will work on my motherboard with that "MANUAL VCORE" thingy?
Please share your knowledge with me!

I would just keep that 600Mhz Katmai and forget about faster CPU. 600Mhz is enough for many old games. If you have games it can't handle consider getting 2nd rig with Athlon XP/P4/Athlon64. Also resell those CPUs that don't work for you, someone may find those PIII 800 useful.

AlexZ wrote on 2021-07-24, 20:28:

I would just keep that 600Mhz Katmai and forget about faster CPU. 600Mhz is enough for many old games. If you have games it can't handle consider getting 2nd rig with Athlon XP/P4/Athlon64. Also resell those CPUs that don't work for you, someone may find those PIII 800 useful.

The reason I want to use this motherboard is because of Voodoo 3 card I recently acquired. Initially I was going to use it on my P4 system but I realized that it does not fit in my 4x AGP slot P4 motherboard.

Most Glide games would be okay with 600MHz but you know, I just wanna give it a little more so that it will not have any bottleneck.

jedikwon wrote on 2021-07-24, 18:04:

One adapter I found shows me "MANUAL VCORE" with jumpers to set voltages from 1.60~2.40.
I don't know if this adapter can really regulates voltages or it just wants me to set correct voltage that written on CPU.

Very likely that adapter allows you to manually specify the voltage that the mainboard should provide to the CPU instead of letting the CPU do that. This is useful if you try to overclock, because you can specify a higher VCore than the CPU requests. But it won't help you getting 1.65V if your mainboard can not generate 1.65V. Still, an adapter like this allows you to set the jumpers to the lowest VCore supported by your mainboard. If the coppermine processor isn't fried by that kind of overvolting, this adapter might make it run in your board.

mkarcher wrote on 2021-07-24, 21:17:

Very likely that adapter allows you to manually specify the voltage that the mainboard should provide to the CPU instead of letting the CPU do that. This is useful if you try to overclock, because you can specify a higher VCore than the CPU requests. But it won't help you getting 1.65V if your mainboard can not generate 1.65V. Still, an adapter like this allows you to set the jumpers to the lowest VCore supported by your mainboard. If the coppermine processor isn't fried by that kind of overvolting, this adapter might make it run in your board.

I see my socket 370 coppermine CPUs have 1.75V written on them. The lowest voltage my motherboard can generate is 1.8V. If I can set the voltage at 1.8V with the adapter and try these 370 CPUs...... profit?!

Do you guys think 370 coppermine CPUs can handle 0.05 higher voltage?

jedikwon wrote on 2021-07-24, 22:03:

I see my socket 370 coppermine CPUs have 1.75V written on them. The lowest voltage my motherboard can generate is 1.8V. If I can set the voltage at 1.8V with the adapter and try these 370 CPUs...... profit?!

Do you guys think 370 coppermine CPUs can handle 0.05 higher voltage?

I don't think these 0.05V are going to kill your processor. The processor will run slightly hotter, and might age faster, but if you don't run the system 24/7, I don't it matters.

jedikwon wrote on 2021-07-24, 22:03:

Do you guys think 370 coppermine CPUs can handle 0.05 higher voltage?

Voltage regulation in those old boards may not be accurate and your CPU may be getting 1.85V instead of 1.75V.

I just checked voltage on my PIII 600 Coppermine in BIOS and even though it's supposed to be 1.65V it kept jumping wildly from 1.4V to 1.8V. I even saw 1.9V spike once. This was happening on all rails for about a minute before CPU voltage stabilized on 1.67V. I would be worried your voltage regulator might kill your CPU this way.

My PSU supplies 12.18V, 3.4V, 5.13V - all quite accurate. But there seems to be a short period until it stabilizes.

jedikwon wrote on 2021-07-24, 22:03:

8><CUT One adapter I found shows me "MANUAL VCORE" with jumpers to set voltages from 1.60~2.40. I don't know if this adapter can […]
Show full quote

8><CUT
One adapter I found shows me "MANUAL VCORE" with jumpers to set voltages from 1.60~2.40.
I don't know if this adapter can really regulates voltages or it just wants me to set correct voltage that written on CPU.
8><CUT

I see my socket 370 coppermine CPUs have 1.75V written on them. The lowest voltage my motherboard can generate is 1.8V. If I can set the voltage at 1.8V with the adapter and try these 370 CPUs...... profit?!

There is one possible 'but'.
There are adapters with voltage regulation below 1.8 volt but that are not able to support s370 Coppermine cpu's because the wiring is only suited for Mendocino Celerons.
Which adapter is it that have you found ?

jedikwon wrote on 2021-07-24, 18:04:

According to this link, Shuttle HOT-661 VRM upgrade , it seems that I need hip6004BCB chip instead of hip6004ACB chip which is c […]
Show full quote

According to this link, Shuttle HOT-661 VRM upgrade , it seems that I need hip6004BCB chip instead of hip6004ACB chip which is currently installed on my motherboard for voltage issue.
Well, sadly, I have no such skill to do that.

And there I found out the existence of this Socket 370 to Slot 1 adapter. It seems promising and I can easily find over 800 MHz Socket 370 CPUs so I want to try it.
However, if the voltage is the problem there is a high chance that the adapter will not work at all.

One adapter I found shows me "MANUAL VCORE" with jumpers to set voltages from 1.60~2.40.
I don't know if this adapter can really regulates voltages or it just wants me to set correct voltage that written on CPU.

The HIP6004ACB only supports down to 1.80V. If the CPU requests a VID for 1.75V or below, the HIP6004ACB failsafes to 0V instead so the CPU is getting no power.
What you suggested would work though. If you use a slocket adapter that has VID jumpers, then you could set it to 1.80V and use a socket-370 chip as you wrote above. You'd have to be careful that it actually supports Coppermines though, and not just Mendocinos.
The jumpers don't regulate the voltage, they just manually override what voltage the CPU Adapter will request from the motherboard, so by changing the jumper it trick the motherboard into seeing a 1.80V CPU which it supports.

I don't think it really matters whether the chip you use was originally rated for 1.75V or something less. Within a given core stepping there shouldn't be any real difference in what range of voltages the chips can tolerate, Intel just specified whatever voltage they thought they needed to get a good yield without being needlessly high. If they thought 1.75V was still safe then 1.80V isn't pushing things much at all.

AlexZ wrote on 2021-07-25, 10:42:

Voltage regulation in those old boards may not be accurate and your CPU may be getting 1.85V instead of 1.75V.

I just checked voltage on my PIII 600 Coppermine in BIOS and even though it's supposed to be 1.65V it kept jumping wildly from 1.4V to 1.8V. I even saw 1.9V spike once. This was happening on all rails for about a minute before CPU voltage stabilized on 1.67V. I would be worried your voltage regulator might kill your CPU this way.

My PSU supplies 12.18V, 3.4V, 5.13V - all quite accurate. But there seems to be a short period until it stabilizes.

It's a slight risk but a very small one IMO.
Onboard motherboard sensors aren't reliable. If the Vcore was really that unstable it should be locking up left and right. If you check with a multimeter you'll probably find it's a lot more stable than what the BIOS is suggesting.
In general, a motherboard's Vcore regulation will be a lot more accurate than the sensors are at reading it.
If that erraticness when cold is really there though, then bad caps could be involved. They work better when they warm up. But the kind of fluctuations that capacitors help with are too quick to see on a meter, it would take an oscilloscope to reliably detect them.

I agree that it's likely to run Vcore a little higher than how the VID is set, but it seems that most boards (or at least retail ones) tend to do that by about 0.05V or so.
I don't know what Intel's specified tolerance is on the Vcore, but I would presume it allows for those 1.75V chips to see 1.80V. Not 1.85V probably, but it's not far off.