Overclock.net › Forums › Case Mods & Cases › Builds & Case Mods › Build Logs › SR-2 Work Station/Gaming Rig build
New Posts  All Forums:Forum Nav:

SR-2 Work Station/Gaming Rig build - Page 7

post #61 of 69
Thread Starter 
Quote:
Originally Posted by ebolamonkey3 View Post
Man, are you getting Hydro EVGA 580s?? That would be suh-weet!
Well that was the latest plan...until yesterday. Now the Nvidia guys have thrown me another curve.

While speaking with one of their Quadro specialists he told me about another technology that they had that uses something called a Tesla board. I had read about their Tesla technology but didn't realize that they had a whole series of boards that were designed around it. I had considered using Quadro boards with a pair of GTX 580's but they said that the two don't play nice together (driver issues). The Quadro guy also concurred that a pair of Quadro 5000 boards would perform better than a single 6000, which also saves over $600 in the process. Then he mentioned the possibility of using a single 5000 with a Tesla board that would be even better yet, and potentially cheaper.

Here's some info on the Tesla board which is basically a giant super computing board that runs with the Quadro card and allows the heavy renders to be accomplished in the back ground while the main card is working on something else.

http://www.nvidia.com/object/product..._M2070_us.html

Check out the 3ds Max iray Rendering Benchmarks.

http://www.nvidia.com/object/quadro-3ds-max.html

The cool part is that I should be able to run one, or more, of these with a GTX implementation since he says that he thinks they are compatible. This may allow me to build something that would still use the new GeForce cards for gaming performance and yet maintain a high level of 3D work station capability at the same time. Sounds too simple huh?

You can't just pick one of these up off of Newegg however so I'm waiting for a call back from one of their Tesla Preferred Providers (sounds expensive already doesn't it). While they couldn't give me any pricing, one of the Tesla specialists did mention that he thought a Tesla card would probably run about the same as a Quadro 5000 ($1700-$2000 range).

So now I'm waiting to see if I might be able to run multiple GTX cards with a Tesla card(s). If so forget the extra card for dedicated PhysX. That will become the Tesla card's new home. Should be interesting. One of the Nvidia guys mentioned that the Communist Chinese now have the world's fastest super computer and that it was basically a giant build of these Tesla boards. What comes after teraflops?
shiva
(13 items)
 
  
CPUMotherboardGraphicsRAM
(2) Xeon 5680 EVGA SR-2 (2) GTX 580's & C2050 Tesla for now 24 gb GSkill 
Hard DriveOptical DriveOSMonitor
Corsair SATA3 256 gb SSD Pioneer B/R & Plextor DVD Windows 7 Pro 64 30" Dell UltraSharp 
KeyboardPowerCaseMouse
Saitek Eclipse (2) Corsair AX1200 Custom Desk Build Logitech G9x & 3Dconnexion Space Explorer 
Mouse Pad
X-Trac Pro 
  hide details  
Reply
shiva
(13 items)
 
  
CPUMotherboardGraphicsRAM
(2) Xeon 5680 EVGA SR-2 (2) GTX 580's & C2050 Tesla for now 24 gb GSkill 
Hard DriveOptical DriveOSMonitor
Corsair SATA3 256 gb SSD Pioneer B/R & Plextor DVD Windows 7 Pro 64 30" Dell UltraSharp 
KeyboardPowerCaseMouse
Saitek Eclipse (2) Corsair AX1200 Custom Desk Build Logitech G9x & 3Dconnexion Space Explorer 
Mouse Pad
X-Trac Pro 
  hide details  
Reply
post #62 of 69
Quote:
Originally Posted by Roark View Post
What comes after teraflops?
it's petaflops and that's actually what they measure the productivity of super computers in. Great build, subbed
SFF Power
(14 items)
 
  
CPUMotherboardGraphicsGraphics
3570k ASUS P8Z77-I DELUXE Intel(R) HD Graphics 4000 NVIDIA GeForce GTX 470 
RAMHard DriveOptical DriveCooling
G.SKILL F3-17000CL9D-8GBSR Western Digital WD6400AAKS Scythe KAZE MASTER Thermalright AXP-100 
OSMonitorKeyboardPower
Windows 7 64bit Asus WV222U Logitech Wireless Solar Keyboard K750 Chieftec GPS-500C 
CaseMouse
Cooler Master Elite 120 Logitech G9x 
  hide details  
Reply
SFF Power
(14 items)
 
  
CPUMotherboardGraphicsGraphics
3570k ASUS P8Z77-I DELUXE Intel(R) HD Graphics 4000 NVIDIA GeForce GTX 470 
RAMHard DriveOptical DriveCooling
G.SKILL F3-17000CL9D-8GBSR Western Digital WD6400AAKS Scythe KAZE MASTER Thermalright AXP-100 
OSMonitorKeyboardPower
Windows 7 64bit Asus WV222U Logitech Wireless Solar Keyboard K750 Chieftec GPS-500C 
CaseMouse
Cooler Master Elite 120 Logitech G9x 
  hide details  
Reply
post #63 of 69
Thread Starter 
Been a while but the UPS and FedEx trucks continue their daily visits. Had to give up on the Hydro's. Every time that they come back in stock they get gobbled up in five minutes and I'm just not that wired in yet to catch it. EVGA really bugs me with that crap but there's more than one way to build a rig.

So I ended up buying a couple of Zotac's and now I'm trying to decide between Danger Den, E.K., Koolance or maybe Aqua for my water blocks. I'm leenin' more toward D.D. right at the moment but they're out of stock on the one I wanted also.

A lot of the stuff that I end up doing has been greatly influenced by folks here in these forums, so I want to thank everyone that contributes 'cuz I've been reading till I go cross eyed, but it's all been a great learning experience. I also spend a lot of time interrogating manufacturers tech people while trying to avoid the sales types. Anyway, thanks again, here's what's doin'.

Got a great deal on the CPU's from ebolamonkey here, thanks again..., picked up the 24 gb of GSkill RAM, 2-Corsair AX1200's, the video cards, M.B., and drives from Newegg, plus a couple of odds and ends from Tiger. Right now I'm trying to finish gathering the water cooling parts while I work on building the desk that this thing is going to live in. The case idea went by the wayside pretty quickly, along with the $8,000 budget I'm sorry to say, but in for a penny in for ten grand I suppose.

The water cooling loop is eventually going to include a geo-thermal ground loop, but for this winter it's gonna be blowing hot air under the family room floor. The system is using most of its plumbing from radiant hydronic heat suppliers. My pump is a bronze TACO mag drive circulating pump. It's 1/25 h.p.. 110 VAC, spins at about 3,200 rpm and should out live me. They have replaceable impeller cartridges but generally they're like the Energizer Bunny. The pump has flanged fittings that go to threaded 1" pipe (CPVC for the present). Flow should be easily adjustable between about 2-6 gpm.

The pump resides in the basement and the line travels up and through the ceiling to the room where the beast will reside. The el shaped workstation/desk that's being built is against only one wall so that in front of the knee hole there is a hidden cabinet that will house the M.B. The RAID array and PSU's are in flanking compartments with their own low velocity ventilation. The area with the M.B. will have a pair of filtered 200 mm Cooler Master 110 cfm fans doing the push pull. The optical drives and the M.B. O.C. jumper panel will be angled in a touch lift compartment built into the desk top. Still trying to come up with a creative way to stow the keyboard and mouse.

The water flow will be controlled by a hydronic system manifold that has integrated flow meters, adjustable balancing valves and isolation valves for each circuit. Each component will have it's own separate supply that I'm going to attempt to run in 1/2" oxygen barrier PEX tubing. It ain't perty but if I can work the bending radius's in you can't beat it for tough and long lasting. Cheap too, except for the fittings, but I've worked with it before.

So once the coolant leaves the components it will all be gathered back into another 1" line that will be fitted to a Tee where it turns to drop back down again. The top of the Tee will have a stack pipe about two feet tall that will stand up against the wall. There will be a boiler sight glass installed on the face of the pipe and a small port Tee will be at the top. The stack is to be left open so that it can function as a combination reservoir, expansion tank, air eliminator, large mouth fill spout, and with the sight glass - a quick check visual assurance that the system hasn't run dry. The small port at the top will be to a small diameter drain line in case of an over fill. I haven't decided whether I'm going to make it copper and try to turn it into some kind of Steam Punk desk, or try to disguise it since that wall also gets some overhead shelves.

Once the line drops back down stairs it will be split and the flow will be divided between two pairs of Thermochill TA 120.4 rads in series. The rads will be using Delta FFB fans (as soon as they come back in stock that is) that can push 150 cfm and make a fair bit of racket (53 dBA). The radiators are being framed together and the rad rack will be hung behind a grill in a wall that separates our basement laundry room from a conditioned crawl space under our upstairs family room. The heat can be directed either way, but the main idea was to try and keep the wind machine somewhere that it won't be heard. In the spring, at least until the ground loop can be finished, the whole thing will fit in the laundry room window so the noise should still be remote enough to be tolerable. From there the water makes its way back to the pump and then back upstairs for some more btu's.

For the immediate, partially because getting a hold of all of the parts has been such a pain and partially due to suggestions made on these forums as well as in conversations with other tech folks, I have changed the video installation for the time being. The initial install is going to be a pair of GTX 580's and a Tesla 2050 card. Maya's supposedly coming forth with new stuff to take advantage of the CUDA cores and it will allow pretty much unlimited editing work while rendering and compiling. They also provide for extremely high precision in CAD/CAM and 3D modeling so that has become a part of the whole now. Once everything is up and running I may add a third 580, but I want to see how everything else gels first.

I had been toying with adding a lift for the 30" Dell so that the whole thing would just look plain Jane till you sat down and then use a pressure switch to have it elevate and turn on all at once. The desk dims would have to change pretty seriously to make it all fit however so that probably won't end up happening. I still need to figure out if Maestro or the TMC will give me the oversight that I desire but the cooling system is going to be set to run on after the system shuts down. I just need to figure out if I should use a timer or thermocouple to control the final shut down.

Also not sure if it's worth the squeeze to try and monitor/throttle the gaggle of rad fans beyond just making sure that they're all running. Still have a lot of investigating and building to go. Thanks again to everyone who has contributed. Whether I follow your advice or not please believe that I do seriously consider all of it and appreciate what you've helped me learn.

As soon as I have something more to show than just piles of parts I'll start posting some pics. Got a new DSLR coming for Festivus, so I'm looking forward to embarrassing myself with some badly lit photography.
shiva
(13 items)
 
  
CPUMotherboardGraphicsRAM
(2) Xeon 5680 EVGA SR-2 (2) GTX 580's & C2050 Tesla for now 24 gb GSkill 
Hard DriveOptical DriveOSMonitor
Corsair SATA3 256 gb SSD Pioneer B/R & Plextor DVD Windows 7 Pro 64 30" Dell UltraSharp 
KeyboardPowerCaseMouse
Saitek Eclipse (2) Corsair AX1200 Custom Desk Build Logitech G9x & 3Dconnexion Space Explorer 
Mouse Pad
X-Trac Pro 
  hide details  
Reply
shiva
(13 items)
 
  
CPUMotherboardGraphicsRAM
(2) Xeon 5680 EVGA SR-2 (2) GTX 580's & C2050 Tesla for now 24 gb GSkill 
Hard DriveOptical DriveOSMonitor
Corsair SATA3 256 gb SSD Pioneer B/R & Plextor DVD Windows 7 Pro 64 30" Dell UltraSharp 
KeyboardPowerCaseMouse
Saitek Eclipse (2) Corsair AX1200 Custom Desk Build Logitech G9x & 3Dconnexion Space Explorer 
Mouse Pad
X-Trac Pro 
  hide details  
Reply
post #64 of 69
Awesome stuff, a couple of things to note:

Don't run that pump at pressures higher than 20 PSI, most watercooling components are only pressure tested to 29 PSI (2 Bar) and you may run into problems with blown o-rings above that. Besides, flow rate doesn't have a very noticable impact on temperatures past 1.0 GPM anyways (Laminar flow stops).

As for the GPU blocks, you should read this amazing review by Skinnee Labs:

http://skinneelabs.com/gtx480-fc.html

It goes over everything from detailed pressure drop testing to the temperatures of VRMs and GPU core. The review was done for the GTX 480 waterblocks but the 580 is so similar that the results will most likely be the same.

If you are going to be using 150 CFM fans the HW-Labs GTX series radiators are a much better deal, here is a graph (purple line is GTX radiator)



And that was with 3k 100 CFM fans! If you are using 150 CFM deltas the gap will be even larger.

Also, I got a good chuckle out of your pressure-switch monitor reveal idea .
Labor of Love
(14 items)
 
  
CPUMotherboardGraphicsRAM
2600k @ 5.0 GHz/1.42v ASUS P8P67 Pro MSI GTX 670 OC 20 GB/1600/7-7-7-24 
Hard DriveHard DriveOptical DriveOS
120GB Vertex SSD Western Digital 1TB None! Ubuntu 12.04/Win8/OSX 10.8.2 
MonitorKeyboardPowerCase
27" 1440p X-Star OCN Mechanical - MX Browns Seasonic X-560 800D w/RX480 top mounted 
Mouse
Logitech USB Generic 
  hide details  
Reply
Labor of Love
(14 items)
 
  
CPUMotherboardGraphicsRAM
2600k @ 5.0 GHz/1.42v ASUS P8P67 Pro MSI GTX 670 OC 20 GB/1600/7-7-7-24 
Hard DriveHard DriveOptical DriveOS
120GB Vertex SSD Western Digital 1TB None! Ubuntu 12.04/Win8/OSX 10.8.2 
MonitorKeyboardPowerCase
27" 1440p X-Star OCN Mechanical - MX Browns Seasonic X-560 800D w/RX480 top mounted 
Mouse
Logitech USB Generic 
  hide details  
Reply
post #65 of 69
Nothing lasts 5 years.Look at what is 5 years old now.
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
I5-750 4.3@1.375v Asus Maximus III Formula Asus ENGTX465 4gb G Skill 2000 Mhz.Trident @2133 9-9-9-27 
Hard DriveOptical DriveCoolingOS
OCZ VERTEX 2 50GB SSD/WD 500GB Blue/WD640GB Black Samsung DVD/RW Corsair H-50 Windows 10 home 64 Bit Retail 
MonitorKeyboardPowerCase
ASUS 24"1080P LED Asus VE248H Abs M1 Cooler master silent pro 600 modular M600 RS-60... Lian Li Lancool PCK-58 
MouseMouse Pad
Steelseries Kinzu no 
  hide details  
Reply
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
I5-750 4.3@1.375v Asus Maximus III Formula Asus ENGTX465 4gb G Skill 2000 Mhz.Trident @2133 9-9-9-27 
Hard DriveOptical DriveCoolingOS
OCZ VERTEX 2 50GB SSD/WD 500GB Blue/WD640GB Black Samsung DVD/RW Corsair H-50 Windows 10 home 64 Bit Retail 
MonitorKeyboardPowerCase
ASUS 24"1080P LED Asus VE248H Abs M1 Cooler master silent pro 600 modular M600 RS-60... Lian Li Lancool PCK-58 
MouseMouse Pad
Steelseries Kinzu no 
  hide details  
Reply
post #66 of 69
Thread Starter 
This is really driving me nuts. I just read a really good rad test the other day and will die a thousand deaths before I can find it again. Looked somewhat similar to the Skinnee report that you showed but more units were tested and the variety of criteria was pretty broad (differing flow rates and air flows with their corresponding temperature drops for example). The results actually showed the Black Ice rads to be quite a poor value over-all. The tests were done just as the new Monsta TFC came out but I honestly can't find the darn link now...but I'll keep looking.

The test that you cite there does not include as many as the one that I had found, and it is actually a comparison with the Thermochill PA model rather than the TA line which has replaced it. The new line is optimized for 130 cfm fans and as such was part of why I intended to use the 150's slightly throttled back, or switch to the EFB's @ 142 cfm or lower and save myself some money for a change.

The phantom test that I can't find included the Magicool Elegant for example but I don't recall the others. Anyway, the flow rates that they tested at were all in the 2 1/4 to 2 1/2 gpm range, rather than the 1 gpm that you mentioned as being the logical limit. Was that conclusion based on water block limitations or radiators?

I had actually begun this hoping to keep the whole thing a lot quieter but quickly realized that I must have gotten a hold of some bad crack if I thought I was going to blow thousands of watts off without an abundance of wind noise.
Faster and louder I'd rather not do. Besides that those higher speed fans suck even more electricity, not that I won't already need my own little reactor out back.

I'm with you on the pressure testing. Remarkably many of the makers advertise that they have pressure tested and yet when you call to ask them they seem to get kind of stupid on the subject. Not callin' anybody out, but several also said that the numbers that they had on their websites were erroneous...not true but false. The pressure's not a problem on mine thankfully, even with the seamingly high head. Static pressure builds 1 pound for every 28" of water column so I'm around 6 before the pump start and right at start I should be somewhere close to 10 at the pump head and maybe a high four on the intake side.

I'd really like to see some testing on the TA series if anybody has seen any yet. I just thought to go with them because Thermochill has been constantly upgrading what was usually considered to be a perennial winner, and then I saw or hallucinated the testing that I mentioned and figure that they would have only made the new one better, otherwise why change it at all. They also did change the threading on their rads with this new series. Now more standard.

Thanks for the GPU test, that's exactly the kind of thing that I've been looking for. I'm going back to read the rest of their tests now.

Curious about your comment regarding laminar flow stopping above about 1 gpm. I'm no hydro dynamics specialist but that figure would be dependent to a large degree on the cross section of the pipe that it is traveling in, as well as the disproportionate differences in the levels of momentum diffusion where the moving water is slipping past the pipe wall that is static. The larger the pipe diameter the lower the number goes. Now I've no real idea how that impacts convection within the pipes conveying the fluid, but once the fluid enters the water blocks, rads or pump they encounter turbulent flow, as they likely would at intersections, Tees, valves and elbows within the pipes. This turbulence has a greater down stream effect in a larger diameter pipe than a small one where the momentum diffusion would tend to help flatten out and realign the laminar layers more readily...I think. Might be completely full of crap and totally wrong...but I don't really think so.

My flow rates are based on 1" supply lines that are having their flow rates halved and quartered when they encounter the supply and radiator manifolds as well. I'm not plumbing a little case with 4' of hose. My loop is the hydraulic equivalent of almost 100' when all of the pressure drops and fittings are considered. Interesting stuff but I'm still a ways from completely understanding all of it. That's what I enjoy most about doing this. I think they used to call it learning.

My next automation is going to be a TEC that cools your CPU while it's boiling wort to brew beer.


Quote:
Originally Posted by charliehorse55 View Post
Awesome stuff, a couple of things to note:

Don't run that pump at pressures higher than 20 PSI, most watercooling components are only pressure tested to 29 PSI (2 Bar) and you may run into problems with blown o-rings above that. Besides, flow rate doesn't have a very noticable impact on temperatures past 1.0 GPM anyways (Laminar flow stops).

As for the GPU blocks, you should read this amazing review by Skinnee Labs:

http://skinneelabs.com/gtx480-fc.html

It goes over everything from detailed pressure drop testing to the temperatures of VRMs and GPU core. The review was done for the GTX 480 waterblocks but the 580 is so similar that the results will most likely be the same.

If you are going to be using 150 CFM fans the HW-Labs GTX series radiators are a much better deal, here is a graph (purple line is GTX radiator)



And that was with 3k 100 CFM fans! If you are using 150 CFM deltas the gap will be even larger.

Also, I got a good chuckle out of your pressure-switch monitor reveal idea .
shiva
(13 items)
 
  
CPUMotherboardGraphicsRAM
(2) Xeon 5680 EVGA SR-2 (2) GTX 580's & C2050 Tesla for now 24 gb GSkill 
Hard DriveOptical DriveOSMonitor
Corsair SATA3 256 gb SSD Pioneer B/R & Plextor DVD Windows 7 Pro 64 30" Dell UltraSharp 
KeyboardPowerCaseMouse
Saitek Eclipse (2) Corsair AX1200 Custom Desk Build Logitech G9x & 3Dconnexion Space Explorer 
Mouse Pad
X-Trac Pro 
  hide details  
Reply
shiva
(13 items)
 
  
CPUMotherboardGraphicsRAM
(2) Xeon 5680 EVGA SR-2 (2) GTX 580's & C2050 Tesla for now 24 gb GSkill 
Hard DriveOptical DriveOSMonitor
Corsair SATA3 256 gb SSD Pioneer B/R & Plextor DVD Windows 7 Pro 64 30" Dell UltraSharp 
KeyboardPowerCaseMouse
Saitek Eclipse (2) Corsair AX1200 Custom Desk Build Logitech G9x & 3Dconnexion Space Explorer 
Mouse Pad
X-Trac Pro 
  hide details  
Reply
post #67 of 69
Quote:
Originally Posted by Roark View Post
This is really driving me nuts. I just read a really good rad test the other day and will die a thousand deaths before I can find it again. Looked somewhat similar to the Skinnee report that you showed but more units were tested and the variety of criteria was pretty broad (differing flow rates and air flows with their corresponding temperature drops for example). The results actually showed the Black Ice rads to be quite a poor value over-all. The tests were done just as the new Monsta TFC came out but I honestly can't find the darn link now...but I'll keep looking.

The test that you cite there does not include as many as the one that I had found, and it is actually a comparison with the Thermochill PA model rather than the TA line which has replaced it. The new line is optimized for 130 cfm fans and as such was part of why I intended to use the 150's slightly throttled back, or switch to the EFB's @ 142 cfm or lower and save myself some money for a change.

The phantom test that I can't find included the Magicool Elegant for example but I don't recall the others. Anyway, the flow rates that they tested at were all in the 2 1/4 to 2 1/2 gpm range, rather than the 1 gpm that you mentioned as being the logical limit. Was that conclusion based on water block limitations or radiators?

I had actually begun this hoping to keep the whole thing a lot quieter but quickly realized that I must have gotten a hold of some bad crack if I thought I was going to blow thousands of watts off without an abundance of wind noise.
Faster and louder I'd rather not do. Besides that those higher speed fans suck even more electricity, not that I won't already need my own little reactor out back.

I'm with you on the pressure testing. Remarkably many of the makers advertise that they have pressure tested and yet when you call to ask them they seem to get kind of stupid on the subject. Not callin' anybody out, but several also said that the numbers that they had on their websites were erroneous...not true but false. The pressure's not a problem on mine thankfully, even with the seamingly high head. Static pressure builds 1 pound for every 28" of water column so I'm around 6 before the pump start and right at start I should be somewhere close to 10 at the pump head and maybe a high four on the intake side.

I'd really like to see some testing on the TA series if anybody has seen any yet. I just thought to go with them because Thermochill has been constantly upgrading what was usually considered to be a perennial winner, and then I saw or hallucinated the testing that I mentioned and figure that they would have only made the new one better, otherwise why change it at all. They also did change the threading on their rads with this new series. Now more standard.

Thanks for the GPU test, that's exactly the kind of thing that I've been looking for. I'm going back to read the rest of their tests now.

Curious about your comment regarding laminar flow stopping above about 1 gpm. I'm no hydro dynamics specialist but that figure would be dependent to a large degree on the cross section of the pipe that it is traveling in, as well as the disproportionate differences in the levels of momentum diffusion where the moving water is slipping past the pipe wall that is static. The larger the pipe diameter the lower the number goes. Now I've no real idea how that impacts convection within the pipes conveying the fluid, but once the fluid enters the water blocks, rads or pump they encounter turbulent flow, as they likely would at intersections, Tees, valves and elbows within the pipes. This turbulence has a greater down stream effect in a larger diameter pipe than a small one where the momentum diffusion would tend to help flatten out and realign the laminar layers more readily...I think. Might be completely full of crap and totally wrong...but I don't really think so.

My flow rates are based on 1" supply lines that are having their flow rates halved and quartered when they encounter the supply and radiator manifolds as well. I'm not plumbing a little case with 4' of hose. My loop is the hydraulic equivalent of almost 100' when all of the pressure drops and fittings are considered. Interesting stuff but I'm still a ways from completely understanding all of it. That's what I enjoy most about doing this. I think they used to call it learning.

My next automation is going to be a TEC that cools your CPU while it's boiling wort to brew beer.
Perhaps they were testing the Black Ice Stealth or X-Flow versions of the HW-Labs radiator in those tests you found. I find it hard to believe that the TA series is "optimized" for 130 CFM fans as it's FPI is so low. A good measure of what fan speed a radiator is designed for is it's FPI (Fins per inch). The TA series radiators have 10 FPI, which is quite low and mainly suitable for low speed fans. The Black Ice GTX radiator has 20 FPI.

The reason the PA/TA series has always been known as the best radiator was because it performed the best across all fan speeds. Sure, it wasn't #1 in any category, but it held it's own across all fan speeds and that was what made it great.

In there testing the reason the testers had 2+ GPM is that they were most likely just running a loop with a pump, radiator, reservoir and aquarium heater. As none of those items are even 1/2 as restrictive as water blocks, their flow would be huge. What I meant is that there isn't a difference in temperature past 1.0 GPM because the heat capacity of water is so high. Below 1.0 GPM you get laminar flow in your radiator and that significantly hurts heat dissipation.

You could run a 2.0 GPM loop, but there would be no real advantage over running it at 1.0 GPM.

You are correct to say that laminar flow is dependent on the size of the passage, 1.0 GPM is just a good safe figure that seems to work for the majority of blocks. Some will be slightly more (wanting 1.2 GPM) and some will be less (say 0.6 GPM). Laminar flow will always be present in your tubing, (which is a good thing, as it reduces restriction). Problems arise when you get laminar flow in your radiators/water blocks.

Basically what you want is:

Tubing: Laminar Flow
Blocks: Turbulent
Radiators: Turbulent
Pump (Inlet) : Laminar
Labor of Love
(14 items)
 
  
CPUMotherboardGraphicsRAM
2600k @ 5.0 GHz/1.42v ASUS P8P67 Pro MSI GTX 670 OC 20 GB/1600/7-7-7-24 
Hard DriveHard DriveOptical DriveOS
120GB Vertex SSD Western Digital 1TB None! Ubuntu 12.04/Win8/OSX 10.8.2 
MonitorKeyboardPowerCase
27" 1440p X-Star OCN Mechanical - MX Browns Seasonic X-560 800D w/RX480 top mounted 
Mouse
Logitech USB Generic 
  hide details  
Reply
Labor of Love
(14 items)
 
  
CPUMotherboardGraphicsRAM
2600k @ 5.0 GHz/1.42v ASUS P8P67 Pro MSI GTX 670 OC 20 GB/1600/7-7-7-24 
Hard DriveHard DriveOptical DriveOS
120GB Vertex SSD Western Digital 1TB None! Ubuntu 12.04/Win8/OSX 10.8.2 
MonitorKeyboardPowerCase
27" 1440p X-Star OCN Mechanical - MX Browns Seasonic X-560 800D w/RX480 top mounted 
Mouse
Logitech USB Generic 
  hide details  
Reply
post #68 of 69
Wow, lots of stuff happened huh? Keep us updated and post a picture once in a while!

Quote:
Originally Posted by Roark View Post
The water cooling loop is eventually going to include a geo-thermal ground loop, but for this winter it's gonna be blowing hot air under the family room floor. The system is using most of its plumbing from radiant hydronic heat suppliers. My pump is a bronze TACO mag drive circulating pump. It's 1/25 h.p.. 110 VAC, spins at about 3,200 rpm and should out live me. They have replaceable impeller cartridges but generally they're like the Energizer Bunny. The pump has flanged fittings that go to threaded 1" pipe (CPVC for the present). Flow should be easily adjustable between about 2-6 gpm.
That's just insane.... but awesome!
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 3570K Asus P8Z77M-Pro Sapphire 7970 Mushkin DDR3 
Hard DriveCoolingCoolingOS
Crucial M4 Thermalright HR-02 Macho Accelero Xtreme Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Dell 3007WFP-HC Deck Legend Ice Tactile Corsair AX650 Silverstone TJ08-E 
MouseMouse Pad
Logitech G500 SteelSeries QcK+ 
  hide details  
Reply
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 3570K Asus P8Z77M-Pro Sapphire 7970 Mushkin DDR3 
Hard DriveCoolingCoolingOS
Crucial M4 Thermalright HR-02 Macho Accelero Xtreme Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Dell 3007WFP-HC Deck Legend Ice Tactile Corsair AX650 Silverstone TJ08-E 
MouseMouse Pad
Logitech G500 SteelSeries QcK+ 
  hide details  
Reply
post #69 of 69
As previously stated by others, I don't foresee a good sr-2 system being too slow in the next few years. =) He should be very grateful and love it. I know I would.
Skoll
(14 items)
 
  
CPUMotherboardGraphicsRAM
5820k@4.4 Gigabyte X99 SOC Champion Sapphire 290X TriX 8GB Crossfire G.Skill Ripjaws V @3200 
Hard DriveCoolingOSMonitor
Samsung 850 Evo Noctua NH-D15 Windows 7  Asus MG279Q 
KeyboardPowerCaseMouse
Custom Poker SeaSonic SS-850 Thermaltake Core X9 Logitech G5 v1 
Mouse PadAudio
Razer Destructor 2 Vali 2 + Modi 2 Uber + ATH-AD1000X 
  hide details  
Reply
Skoll
(14 items)
 
  
CPUMotherboardGraphicsRAM
5820k@4.4 Gigabyte X99 SOC Champion Sapphire 290X TriX 8GB Crossfire G.Skill Ripjaws V @3200 
Hard DriveCoolingOSMonitor
Samsung 850 Evo Noctua NH-D15 Windows 7  Asus MG279Q 
KeyboardPowerCaseMouse
Custom Poker SeaSonic SS-850 Thermaltake Core X9 Logitech G5 v1 
Mouse PadAudio
Razer Destructor 2 Vali 2 + Modi 2 Uber + ATH-AD1000X 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Build Logs
Overclock.net › Forums › Case Mods & Cases › Builds & Case Mods › Build Logs › SR-2 Work Station/Gaming Rig build