VOGONS


First post, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I bought four 400Mhz Celerons which all overclocked easily to 500Mhz with an 83.3Mhz bus, but the two 533Mhz ones I bought all failed to boot. Has anyone had any success with getting a 533Mhz Celeron clocked up to 666Mhz with an 83.3Mhz bus? I'm wondering how difficult it is to win this silicon lottery and if it's worth the expense or trouble.

I am trying to max out my 440LX. Do I have any other options from other non-Intel compatible processors?

Reply 1 of 6, by BitWrangler

User metadata
Rank l33t++
Rank
l33t++

From what I remember, they don't go very far over 600 without extreme measures, have heard of 670, but that was regarded as unusually high/lucky. Think it took voltage and watercooling to do it.

The 83mhz thing, I never understood back in the day why it was more of a problem on 370/slot1 than on socket 7... data corruption wise. Though reflecting on it lately, I think it might have been due to more of them being built by ppl who would get the new bleeding edge DMA33 HDDs, while socket 7 penny pinchers would stick with DMA 2 8GBs they already had or got cheap. Also socket 7 penny pinchers would use up old sinks on the north and southbridges, if it's warm it gets a sink, intel guys were more like "unless there's one made specially for it, it doesn't need one" and trusted those useless quarter inch high things that were for bling more than cooling. Many factors though.

Anyhoo, I'd say if you want to do your best to avoid it, overspec your drives and cables or lock them down in setup to a lower mode if possible, keep everything as cool as you can get it, and don't have anything irreplaceable on the drive.

Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.

Reply 2 of 6, by Kahenraz

User metadata
Rank l33t
Rank
l33t

If I overclock the FSB, should I put a heatsink on the chipset? It's not scalding to the touch, but it does get too hot to keep my finger on it for too long. This never ocurred to me before, as these chipets have always been bare. I don't think it will make a difference whether a 533Mhz chip will POST from a cold start, but maybe it will help with stability?

Reply 3 of 6, by BitWrangler

User metadata
Rank l33t++
Rank
l33t++

Yah if it won't even POST then you are not gonna change that with the sinks, might keep it running after you get the POST but won't help get it. Did you try slackening off the timings as far as they go while at normal speed, saving that then going to 83? And how is the RAM, if you've got some low end barely PC66 stuff in there it might be the stopper.

IMO no point bumping the core voltage to get a POST, it's only really useful to get stability at the highest you can POST rather than giving you anything extra. Diminishing returns on it too, just helps force the last 20-50mhz.

Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.

Reply 4 of 6, by Kahenraz

User metadata
Rank l33t
Rank
l33t

When you say slackening the timings, you mean the memory, right? I'm using a single stick of 128MB PC100 memory.

Reply 5 of 6, by dionb

User metadata
Rank l33t++
Rank
l33t++

Which Celeron 533 are we talking about here? The fastest Mendocino and slowest Coppermine Celeron ran at this speed. The Mendocino is right up against the limits of its design, I'd not expect you to be able to hit 75MHz let alone 83 with that. The Coppermine is the exact opposite - they frequently ran up to 100MHz and on an LX board the chipset will be the bottleneck, not the CPU.

Reply 6 of 6, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I'm referring to Mendocino.