What does /cemetech/ think of mineral oil computers?

If not mineral oil, what oil would be betteR?

It always seems so shady that it would work without gumming up the works and shorting everything, but from what I've seen on the 'net it works fine.

mineral oil is way better than vegetable oil. I would love to try it, not because it necessarily cools better (as it doesn't), but because it just looks friggin awesome. There really isn't a better oil that I know of, as mineral oil is pretty ideal - mostly clear, non conductive, etc...

I assume you've seen this? http://www.pugetsystems.com/submerged.php

Yeah, I saw that a while ago and I am really wondering why it does not crash at those temperatures... (It goes over 80 C with load, so why not? What can the oil do to it?)

EDIT: I looked at it again and I guess they updated it since I saw it =)

80c is still within "spec" for CPUs - it was pushing the upper limits, but it did not exceed them. Overclocking at those temps is unlikely, but at stock it should (and does) run just fine.

However, as they noted, lifespan is likely to be significantly reduced running at such high temperatures, hence the giant radiator which dropped load temps to 45c - not bad at all, not bad at all...

why doesn't the oil ruin the motherboard?
i don't get it... wouldn't it short it out or something?

the oil does not conduct electricity, so it does not short out the board.

that's really cool! that would be sweet to have
in my point of view at least...

The whole idea of oil submerged computers is great in theory. However, a standard motherboard cannot be used under oil where the oil is in direct contact with the board. The oil gets into connectors, and, since it is an insulator, stops data transfer. So you could submerge a board under oil, but expect the ram to stop working fairly quickly. The ram is not damaged, just not making good connection with the board since it is covered in mineral oil. Other connections, such as power and the bios battery would also suffer from the same problem. One possible solution to oil on the connectors is to cover all connectors with a silicone sealer caulk, but this would make it hard to upgrade components of the system. Naturally, you cannot use most harddrives under oil, since the harddrives are vented. The oil will slowly seep in, and add resistance to the head and disk, and possibly causing some abrasion. This sounds like a cool project to undertake though, if you use some planning before you soak your board in oil. I might try it with one of those old pIII boxes sitting in my bedroom .

You can use an filter designed for aquariums to pump the oil through a heatsink assembly to cool the oil. I am thinking about building this. I have peltier modules I can use to drop the oil temp to around -5 degrees C. This may work. If I undertake this project, of course I will post photos.

Take a look at the link he posted, it works fine.

Spoken like a true fool. "In theory it will work, but in practice it totally won't. I refuse to acknowledge direct evidence and pictures from a reputable company stating/proving otherwise" Oh, and don't double post - that is what the little button with the text "EDIT" is for...

Anyway, after two months it was still running fine with load temps of 45C due to the addition of a radiator and pump. The site says a 6 month update should be posted soon, and seeing as they haven't ripped the site down and replaced it with a "please don't try this" message, I'm guessing that it is still working just fine

Besides, your theory as to why it wouldn't work has serious flaws with it. For one, it can't "get in the connectors" as it was assembled while still dry. The connections (especially RAM) are quite tight and there is no way that oil would be able to force its way in there and eliminate all metal-on-metal contact. As for upgrading, sure, that is a (somewhat) valid point, but it isn't all that difficult to shutdown the system and simply pull it out of the oil before upgrading it (cleaning it first). The hard drive wasn't submerged in the oil, by the way.

THat is a very awesome computer An once the radiator was added you could do almost unlimited gaming without it over heating. it is one of the coolest looking computers I have ever seen.

Reread the post you're replying to. The proposed upgrading issue was in regard to using sealant to keep oil out of the connectors, not in regard to the oil surrounding the components.

Correction: You *can* do unlimited gaming, as it won't overheat since it has a load temp of 45c. However, given as they have it running 100% load all the time, I'm guessing they have something like folding@home running on it.

Just because one entity built an oil submerged computer, it dosen't mean that all computers can be used in this way, or even how long they will work. While oil cooled systems are a great idea, more testing needs to be done to determine their long term durability, and the ability of all computers to be cooled in this manner.

Yes, because computers vary wildly in their construction and design

Oh, and they have submerged two computers in this fashion (one was some old POS to test that it wouldn't fry the computer), and many have done vegetable oil submerged computers just fine. Six months of the computer in mineral oil is a pretty good indication that everything is just fine. Don't get all pissed off because your statement/theory got smashed to pieces before you even made it and destroyed any chance of you having a reputation for providing correct answers.

Here is the vegetable oil submerged computer: http://www.tomshardware.com/2006/01/09/strip_out_the_fans/

So, to be quite honest, I've been informed I am not allowed to purchase a 5 gallon drum of mineral oil

According to the link posted above to the page with the vegetable oil cooled system, sealing connections is necessary. "The CPU base must be sealed tight against seeping oil - otherwise there will be problems."-http://www.tomshardware.com/2006/01/09/strip_out_the_fans/page8.html
In order for an oil-cooled system to work, according to this page you must seal around the processor with silicone, although not for the purpose of the oil acting as a dielectric and preventing contact at the leads of the processor.

Then perhaps THG was wrong (which, frankly, happens a LOT - that site sucks balls) and just playing it safe, or mineral oil doesn't require such measures.