Threads in here keep getting closed, why the censorship?

You can still read them, so no censorship. ;-)
This is to stop people from bumping ancient threads, useful on active forums, here locking threads automatically after 6 months is of course completely useless.

I have 6 of these pc's and they each have dual Intel Xeon quad core cpu's that think they are 8 core cpu's

so 16 cpu cores in each pc AND I got them for $250US shipped to my door only needing a harddrive and a gpu to run.

They all came with 16gb of ram as well which I have upgraded to the pc max of 32gb except 1 pc that has 2 bad memory slots in it so it only has 24gb of ram in it. They crunch everything I've thrown at it so far except LHC and LHCdev but that's probably something in the software, Linux or Windows, that's not been updated.

I have used an adapter in regular Dell pc's before to upgrade the psu and that has worked great, they make 24pin to 8 pin Dell mb connectors. The only thing is they have to stay in the original case because of the way the power on/off, the sound and the restart buttons are wired into the mb they use a plug not individual wires and I've never found an adapter for that and I'm not running a bump switch to start the pc.

The older pc cases have very poor airflow and no top fans at all so modern day gpu's don't do well in them.

When they're not powerful enough, you don't need a bigger compatible one, you just need more on the 12V line. Just get any old PSU and use it in addition.

Works only with PSUs, that don't require significant load on 5V to keep the 12V line at the right voltage. Some won't even turn on without something @5V.

By the way metal shelving is a bad idea, you're asking for shorts. I use wood.

I understand why you are worried about shorts, but in general, as computer cases are made of metal, there's nothing wrong about putting them on metal shelves. Server racks aren't made of wood either and nobody worries about shorts there eiter.

Because my pc's are on wire shelves like this

Probably you even can't use anything else than the original PSUs, the HP mainboard seems to have non-standard 18-pin connector according to the pictures I see on Google and Dell is known to use own wiring on standard ATX connectors, so you'll burn the mainboard and/or the PSU if you use anyhing else then the original PSU.

I'd never buy a desktop computer not build out of standard parts, when anything fails, it's just scrap as replacement parts are usually very expensive for such systems or not available at all once they run out of support. Well, I'd never buy a prebuild desktop computer, but that's another thing.

I have an Nvidia Quadro K600 (981MB) gpu on it and it crunches Moo just fine. I can't put a more powerful gpu in it because the PSU is too small and it's one of those case length ones that HP and Dell used. They are WAAAY too expensive to replace and there's no room for an external psu. The pc also has 2 quad core Xeon cpu's in there as well, so 16 cpu cores.

I believe processing numbers _2 and higher helps clear the backlog. It doesn't necessarily mean the validator is crashed it just means it's struggling to keep up with demand. In the past I believe this has been partly to do with the fact the ready to send queue has been so large. I believe admins had an idea how to fix it and it would appear that has worked

That is the `problem` of having a room full of parts ,
every now and then they add up to "if I put that bit and that lot in that case if should work"
so long as I buy that bit from ebay . . . .
it`s a never ending story of not wanting to chuck stuff out until its dead :-)
And the amount of "dead" or "For parts only" that sell on ebay is incredible .

Perhaps I could increase the cache next time I see waiting for validation piling up.

When I see the validation queue starting to pile up this is when I will start contributing in clearing any _1 or higher numbers before processing any _0 tasks. I haven't seen a pileup in quite some time

Doesn't Milkyway have stuff available 99% of the time?

Yes, but sometimes it doesn't, I'm not running huge cache.

Current progress.

As of today, "The odds are 1 in 20,842 that we will wrap this thing up in the next 24 hours. (This also means that we'll hit 100% in 20,842 days at yesterday's rate.)"
As of 2023-04-23 at 23:59 UTC 10.000% of the project has been completed. Thanks to everybody that helped get to this point

+1
Good idea I have been wondering how we are getting along with this .
I did read somewhere [some internet poo site] that it `could` take `up too` 150 years to crack one of thease
though that's probably only if I use my old AMD K6-2 550 (it is overclocked to 604 MHz :-) it was fast in its day . . . .

Yes, I agree, a direct link to the current progress either from the "project" menu or a "science" menu (which is missing/disabled here) would be nice to have.