the celeron should be a bit faster then the pentium 2 333. An option could be trying to find a slot1 adapter for the PPGA celerons. They are the same core as your celeron 433 but they go up to 533Mhz.

If the motherboard supports it, overclocking might be an option as well, I have never seen a mendocino celeron that wouldn't do at least 75Mhz, most tolerating 83 as well, but YMMV.

On Tuesday we might have the system. I have considered overclocking to 500mhz. I hope that motherboard has some overclocking options.

As a user of the said combination, it should be all right. I benchmarked my 433 MHz Celeron with NSSI and the results it produced were similar to a 400 MHz Pentium II (rough estimate), those Celerons do have a place in motherboards that are limited to 66 MHz FSB CPUs (like the older 440LX boards which is what you're dealing with).

Mendocino Celerons and Deschutes P2s have extremely similar clock-for-clock performance, with subtle differences due to the Mendocino having faster but smaller L2 cache and the Descutes larger but slower cache. The only advantage a P2-400 would have is the 100MHz FSB, but with an i440LX board that's not an option. A Deschutes 333MHz would perform almost identically to the Mendocino Celeron 333A, a Celeron 433 would be significantly faster.

https://www.anandtech.com/show/274/5

Anand's testing is limited to Quake 2, but it gives an idea of where a Celeron 433 falls against other Slot 1 CPU's in a period-correct game with a Voodoo2. The Celeron 433 manages to punch up between the Pentium II 400 and 450. It's definitely going to be faster than a Pentium II 333 in pretty much any game.

dionb wrote:

Mendocino Celerons and Deschutes P2s have extremely similar clock-for-clock performance, with subtle differences due to the Mendocino having faster but smaller L2 cache and the Descutes larger but slower cache. The only advantage a P2-400 would have is the 100MHz FSB, but with an i440LX board that's not an option. A Deschutes 333MHz would perform almost identically to the Mendocino Celeron 333A, a Celeron 433 would be significantly faster.

Another advantage that at least some of the 400MHz Deschutes CPUs have, is an unlocked multiplier (mostly downwards).
While Mendocino was locked, these Deschutes can be set to 66MHz FSB and upping the multiplier I got it booting at 366MHz.
Just wanted to add this small bit of information 😀

Unfortunately these seem to be very rare while celerons always remained cheap.

On the other hand, I wonder if one can't mod a bios/slotcket to insert Coppermine celeron into a mobo with i440LX chipset. These went all the way up to 766MHz so that would make a nice boost !

I had a Celeron 433 (slot 1) back in the day. I ran it at a 75MHz bus speed that brought it up close to 500MHz without any issues. 83MHz would boot but wasn't reliable in Windows. I had a Voodoo3 3000 AGP though and not a PCI Voodoo 2, so YMMV if you decide to go the old school OC route.

I remember Intel keeping Celerons on a 66MHz FSB for the longest time. On a 100MHz FSB they were too competitive with the P-II's they were trying to sell at the same time. Most people would get a Celeron 333 and it would run at 500MHz with 100MHz FSB without problems.

A few days have passed since my buddy bought the retro-pc, I've set it up with win98se on a 15gb IDE WD Protege, and installed the Voodoo2 and everything seems fine but now I need your help in order to upgrade the cpu. The system has a Pentium2 266mhz with the classic black plastic adapter. I don't know if it is possible to open up the adapter in order to replace the cpu, without breaking it. The Celeron 433 I have comes with its own adapter "370 cpu card" from a QDI board (with VIA chipset). I can tell that this "370 cpu card" fits perfectly on the P2L97-S's slot. But after installing the new cpu the system doesn't seem to work. It boots but with no signal on the monitor. So I guessed that I need to update the bios. It seems that the board right now has the oldest bios available, the 1001 version. I downloaded the latest version 1010, also found the tool called pflash.exe and made a startup floppy disk after I followed the instructions from the manual. After booting from the diskette and running the pflash tool I chose "Update bios including boot block and escd" and then typed the actual name of the bios file. Everything seems ok, but when I hit enter after the "Are you sure?" message in order to start the actual update, I get an Error "Unknown Flash Type" message. What do I have to do in order to update the bios? Do I have to select in the Advanced options the first option "Clear PNP ESCD Parameter Block" first? I'm afraid that might corrupt the current bios... Help please!

The whole black plastic unit of the pentium 2 IS the CPU, you cannot remove any chips in there. Except for the pentium II overdrive, none of the pentium 2s were made for sockets. The main chip is soldered on the cartridge ant it won't work either without the L2 cache soldered nearby.

As for the bios, are you sure that's the right one ? If you are, then try uniflash. It will flash whatever rom you want to whatever mobo you chose.

For more information, research the Slot 1 form factor. Additionally, the "370 CPU Card" is known as a "Slotket", a portmanteau of "Slot" and "Socket" which allows a socketed processor to be installed into a Slot based system. The are a number of these that may be physically compatible, but electrically incompatible. This shouldn't be a problem in this case, but arises with Coppermine and Tualatin CPUs due to the differences between "PPGA370", "FC-PGA 370", and "FC-PGA2 370".

Confirm that you are using a heatsink of some sort on the Celeron, and that you have altered the jumpers appropriately, according to the manual.

Thanks for your reply. So it might not need a bios update? This celeron is a Mendocino and the "slocket" is a PGA370. Here is a picture of it:

(Ofcource I also have a heatsink, just removed it)
What jumpers do I have to alter in order to make it compatible? If you mean the motherboard's the only jumpers I can alter are the multiplier's but it seems that this board only supports up to 5.0x multiplier and the celeron has a 66mhz bus speed with a 6.5x multiplier.

Deksor wrote:

The whole black plastic unit of the pentium 2 IS the CPU, you cannot remove any chips in there. Except for the pentium II overdrive, none of the pentium 2s were made for sockets. The main chip is soldered on the cartridge ant it won't work either without the L2 cache soldered nearby.

As for the bios, are you sure that's the right one ? If you are, then try uniflash. It will flash whatever rom you want to whatever mobo you chose.

Thanks for the valuable info. The bios is the correct one, downloaded from the board's page https://www.asus.com/supportonly/P2L97-S/HelpDesk_BIOS/. Another thing that bugs me is that the original bios file name ends with .awd (LX2S1001.AWD) whereas the latest bios file is 1001S.004. Should I rename the ending to "awd"? The uniflash tool is compatible with this board? And if so What version should I download?

Without clear identification of the card or jumper documentation we can't say how the various jumpers need to be set. Is there anything written on the back of the card that could help identify it?

That said:
1) you don't need to worry about the multiplier as all Celeron Mendocino chips (other than Engineering Samples) have locked multipliers. So it doesn't matter what you set, it will be ignored anyway.
2) if that was a working pull from the QDI board, the jumpers should be set correctly for this CPU anyway.

Nonetheless voltage and bus speed settings (the other two jumper blocks) may not be correct or ideal, so it's worth checking.

However my guess is that the BIOS simply does not understand Celeron Mendocino CPUs and requires an update before the board will boot with them. As for the Unknown Flash Type error - does it look like the BIOS EEPROM (flash) chip has been messed around with (signs of stickers being torn off or maybe even stickers for another board on it)? In that case there could well be a different chip there than the flash program expects. I for one am very pragmatic/lax about selection of EEPROM chips if I need one for a board: as long as it's electrically compatible and the right size, I use it. I never (or rather: rarely) hit these unknown flash errors as I use Uniflash. If that board comes from someone like me, that could explain the odd flash chip. Using uniflash would solve your problem. Then again, if it's running the very first 1001 BIOS, it sounds pretty stock - I can't imagine why anyone would downgrade to such an old version... more info needed...

Edit:
I just checked the manual - it refers to pflash2.exe being the correct tool to update the BIOS, not awdflash.exe. Possibly that's why awdflash doesn't recognise the chip.

Edit2:
Uniflash 1.40 is the version to use (I've tried a newer patched 1.47 and it failed spectacularly, so stick with the old one - it's still a decade younger than your board), available all over the place, e.g. here: http://www.rainbow-software.org/uniflash/

dionb wrote:

Without clear identification of the card or jumper documentation we can't say how the various jumpers need to be set. Is there a […]
Show full quote

Without clear identification of the card or jumper documentation we can't say how the various jumpers need to be set. Is there anything written on the back of the card that could help identify it?

That said:
1) you don't need to worry about the multiplier as all Celeron Mendocino chips (other than Engineering Samples) have locked multipliers. So it doesn't matter what you set, it will be ignored anyway.
2) if that was a working pull from the QDI board, the jumpers should be set correctly for this CPU anyway.

Nonetheless voltage and bus speed settings (the other two jumper blocks) may not be correct or ideal, so it's worth checking.

However my guess is that the BIOS simply does not understand Celeron Mendocino CPUs and requires an update before the board will boot with them. As for the Unknown Flash Type error - does it look like the BIOS EEPROM (flash) chip has been messed around with (signs of stickers being torn off or maybe even stickers for another board on it)? In that case there could well be a different chip there than the flash program expects. I for one am very pragmatic/lax about selection of EEPROM chips if I need one for a board: as long as it's electrically compatible and the right size, I use it. I never (or rather: rarely) hit these unknown flash errors as I use Uniflash. If that board comes from someone like me, that could explain the odd flash chip. Using uniflash would solve your problem. Then again, if it's running the very first 1001 BIOS, it sounds pretty stock - I can't imagine why anyone would downgrade to such an old version... more info needed...

Edit:
I just checked the manual - it refers to pflash2.exe being the correct tool to update the BIOS, not awdflash.exe. Possibly that's why awdflash doesn't recognise the chip.

Edit2:
Uniflash 1.40 is the version to use (I've tried a newer patched 1.47 and it failed spectacularly, so stick with the old one - it's still a decade younger than your board), available all over the place, e.g. here: http://www.rainbow-software.org/uniflash/

The other motherboard is the QDI Advance5-133E and is in working condition. The BIOS chip is intact, with its sticker on. I cannot find any other jumpers for the voltage control on the asus board apart from the bus speed and multiplier. There is only a "RCPU" jumper near the cpu slot which is open.
edit: I used the pflash.exe not the pflash2.exe. I'll try this first.

Then that's your problem. Check the relevant pages of the manual, it not only tells you you need pflash2, but also why...

dionb wrote:

Then that's your problem. Check the relevant pages of the manual, it not only tells you you need pflash2, but also why...

I was looking at the wrong manual! However you also might look the wrong one too. This manual https://www.asus.com/supportonly/P2L97-S/HelpDesk_Manual/ refers to AFLASH so I'll use that instead.

I was looking at the original 1997 manual, you have a 1999 update. Shouldn't make too much difference:
http://oldschooldaw.com/asus/1997/p2l97s-105_1997.pdf

Either way, if you're sure you have the right BIOS file, just use Uniflash. If you have anther board working with the same type of BIOS chip (32p EEPROM), the worst that can happen is that you totally misflash the P2L97-S, in which case you can recover by hot-flashing the chip on that board.

I did the job with AFLASH as the manual recommended and now the celeron works perfectly! I'll do some benchmarks and post the results here. Thanks everyone!