He's right about the pump issue. I'm not too familiar with ups and downs on pump but every time your tubes goes up and down and in and out of radiators it's fighting head pressure?
My XSPC is rated at 1.8 m of head pressure but just about 3 feet of line going up and down from rad to res. already got it crawling back at a minimum flow. Most pumps don't do more than 10m of head pressure and even if you get good monster pumps remember your GPU/CPU block max flow rate. I mean you could have 50 feet of head pressure and 30 gallons per minute pumps but if your cpu block like mine a XSPC block limits flow to 1 gallon per minute, yea u got a bottleneck lol. I guess that's where multi-GPU/CPU setups come in handy so that flow can be used elsewhere.
Yeah I've seen that video. I can understand wanting to make it as small as possible, but to then later complain about the resulting heat output and throttling is a little silly.
i think it's such a waste since they are trying to move into a new office space and when they do they'll soon have to tear all that shit down.
I am pretty sure that they started this before they decided to move.
Yes, I am fairly sure of that too, I do think it was like 4 months ago or so, they started on this project.
I'm about 99% certain that Linus said the rads on each PC were there only if the systems had to be removed from the master loop and run standalone. The idea is to have a connection hanging there at each station that the PC's loop can be directly attached to. IF you need to take it somewhere you disconnect from the "room" and hook the pc's tubes into the top mounted rad. This setup is actually very smart and does NOT use the individual rads when on the "room" rads.
Now about the copper piping, they should definitely insulate it or it will be a lot less effective.
I recently watched the video where they are asking for money (We are moving or something like that). You can see the copper piping in the backgroud at the end, and you can see what looks to be hoses running outside of the bathroom window (even though they drilled holes in the wall opposite the bathroom). Check it out and see if oyu notice anything that might hint at what they are actually doing if you are interested.
There is also resistance in the piping which increases pressure drop (i.e. required head) of the loop. pipe (or tube in your case) resistance is proportional to length and inversely proportional to diameter.
I think they will have to use several pumps throughout their loop to maintain flow without over pressurizing the water blocks.
I like your name.
Do you think you could link to that video?
https://www.youtube.com/watch?v=_cLAmvLmzVE
https://www.youtube.com/watch?v=_cLAmvLmzVE&list=UUXuqSBlHAE6Xw-yeJA0Tunw&feature=player_detailpage#t=84
Point of interest is ~ 1:20
Pipes are still bare and judging by the state of the room this is the finished state. I have no idea what the tubes running through the bathroom window are, it may well lead to the outside radiator.
I guess we'll have to watch his videos to find out more interesting questions, but what I am currently most interested in are: 1) what is their pump configuration? They will need to maintain flow throughout the loop without overpressurizing any components, so they probably have to use several relatively low pressure pumps in series throughout the loop. I am thinking at least one on each side of the large radiator plus at least one for every two or three computers on the loop. 2) Where is the radiator and can they guarantee it won't freeze during the winter? 3) Do they have any kind of condensation protection for their components for when the outside temperature is significantly lower than the inside temperature?
It looks like one of my criticisms (radiators mounted to the cases) may have been unfounded. After reading some comments here and rewatching the video, I think they may be using some quick disconnects so that they can switch between either whole room water cooling or local water cooling on the fly.
im sorry, i didnt see you guys in the video.
Personally, i have never done any custom watercooling ( of a PC anyway ) and have absolutely no official training or education with regards to physics etc.... so feel free to correct me if am wrong, as everything i am about to say is based purely on my experiences whilst pissing about in my garage, mostly with cars...
But with your second question, assuming the radiator is outside, and temps outside do occasionally hit freezing level.... assuming they got hold of an older style ( copper pipe+copper fin ) car rad, and had it properly cleaned... could they not ( in theory anyway ) just use an industrial/ automotive anti-freeze solution diluted with ph neutral water ( or coolant of their choice ) in order to eradicate the freezing issue??
Also, i believe this would deal with the majority of galvanic corrosion ( aswell as general corrosion ) due to the setup being ( almost ) entirely comprised of copper ( Copper rad ~> copper pipes ~> copper cpu/gpu blocks ) along with the majority of these anti-freeze solutions also containing anti-corrosion chemicals...
Just a thought... though i could be completely wrong :-)
Antifreeze would be necessary since they are in Canada. As far as I am seeing things, what they are doing is closer to a car's cooling set up than it is to a typical pc watercooling set up, so you are pretty much right on the money as far as I am concerned.
The big issue that that raises though is the condensation that it would cause. If it is colder outside than it is inside, then the water (+antifreeze) would likely drop below room temp while in the radiator (though it might not when all of the computers are maxed out). When the water comes back into the room, especially in the copper pipes, it would cause condensation. Condensation and computer parts don't mix. They need to address that some how.
Arent they planning on moving to some new place anyway? Seems like lots of work when they said somethinga bout a new office?
Actually having the bare copper pipes means that there will be minimal temperature shifts between ambient and computer, and thus there should not be condensation on the computer components which could be a problem.
The condensation would occur when the water is cooled by a radiator exposed the temperatures far below the ambient temperature indoors. The bare copper pipes may warm up the cold coolant a few degrees, but in the dead of winter, I seriously doubt it will be sufficient to warm up the coolant and cool down the room enough so that the temperature difference does not cause condensation.
Part 2 is up, and I have concluded that they have NO idea what they are doing when it comes to building stuff. I am glad that Linus' (or is it Luke's?) dad is there to give them a hand.
https://www.youtube.com/watch?v=L6GVEx916ik&list=UUXuqSBlHAE6Xw-yeJA0Tunw