Sign in with
Sign up | Sign in
Your question

Need help with choosing WC components

Last response: in Overclocking
Share
April 17, 2012 7:23:57 AM

I'm planning to build a new machine and I'm thinking of going WC for the noise (primary) and the temperature (secondary) reasons.
It's going to be my first WC build, I've been reading reviews for about a month already but I still fill a bit lost :/ 

So the hardware I'm planning to get:
cpu: i7 3820k
mb: Rampage IV (formula or extreme)
graphics: one or two GTX 680 (as soon as they become available)
power: Corsair AX850
case: ?

WC components:
pump: Swiftech MCP35X
cpu block: Koolance CPU-370
rad: Aquacomputer Airplex Revolution (triple)
res: Swiftech MCRES Micro Rev. 2

So the questions:
- Is it a sane WC configuration? Is it an overkill?
- I get that it should be more than enough to cool the CPU only, but what about CPU+one GPU? CPU+two GPUs? Do I need one more rad? (skinneelabs says it can dissipate 600w)
- What case would be a good fit? I've read some good reviews on Obsidian 800D... but it seems folks have different opinions on that..

More about : choosing components

a b à CPUs
a c 78 K Overclocking
April 17, 2012 9:33:07 AM

cases - re all upon users preference's. Through modding, just about anything can go on/in a case. no modding, want to keep it vanilla? well choices become limtied or really bulky!

pump - good
rad - thas an old rad which is also a bit...erm tests show that there are better performing rads for that price.
Quote:
skinneelabs says it can dissipate 600w
have you seen what the fan RPM range was when dissipating 600W's?
Quote:
should be more than enough to cool the CPU only, but what about CPU+one GPU? CPU+two GPUs?
the watercooling sticky shows you how to do a TDP calc and measure what sort of heat load you're looking at. Thus the choice or raddage will be decided upon.

res: Swiftech MCRES Micro Rev. 2 - you sure? thas an old res and there are better available , but upto you

block - XSPC raystorm is a better performing block for cheaper.
m
0
l
April 17, 2012 9:58:21 AM

I would go with mpc 655 pump the apogee hd water block an xspc double pass 120 x 480 any rez will do and place it in an tj07 silverstone all internal.. i built one and did vid on youtube under robustus64 check it out.. its cooling 2 x 480 and a 3930 @ 4,8 ghz cpu never gets above 61c gaming..
m
0
l
Related resources
April 18, 2012 8:45:14 PM

I really can't speak for the pump performance. I have always used D5 pumps so I haven't the necessary experience with the product to make a suggestion.

As for the block, its a good choice. But also have a look at the Raystorm (as stated above),Supreme HF and DD M6 series of blocks.

As for the RAD, I would go with an RX, XTX, GTX or the SR-1 series of RAD.

The Res, I have also never used.

As for your chassis, it all comes down to your budget.

At under 200USD: I would suggest the Switch 810. Never used one, but I've heard nice things.

At under 300USD: I would indeed go with the 800D. Though I do think it is overpriced given the materials (metals/paints) used aren't really the highest of quality.

At 300USD: I would go with one of the many Mountain Mods CYO cases. Customization is king with mountain mods and you would almost certainly find what you are looking for there.

At 400USD: I would highly suggest the CaseLabs M8. I use the M8, and it may very well be the last case I ever need.
m
0
l
a b à CPUs
a b K Overclocking
April 18, 2012 9:22:45 PM

The 35X would be my pump of choice. It has a higher pressure rating so you'll be able to run more components with better flow. The MCP655 isn't bad (using one as I write) but considering you can pick up a 35X from Jab-Tech for the same price I paid for my 655 Vario, it's a steal.

I feel like a single rad won't be enough, especially since you want to cut back on noise. You're looking at ~600W of heat generation in the loop, so I would look for 2 good 360 rads that can dissipate 300+W with quieter fans (~2000 RPM). The XSPC RX rads are very good in that area since they have a lower FPI count.

I second the MountainMods and Caselabs cases. They look fantastic and can really be built to you specs/needs/wants.
m
0
l
a b à CPUs
a c 78 K Overclocking
April 19, 2012 2:27:32 AM

there's also the Alphacool UT60 360mm full copper rad thas a well performing all rounder from low~upto high rpm ranges.
m
0
l
April 19, 2012 8:58:45 AM

Thanks everyone for your replies. Let me take another round of suggestions and questions...

Ok, I read some more reviews, browsed a couple of different products.
What about:
Apogee HD vs XSPC RayStorm?
HWLabs GTX 360 vs MCR-X20 Drive Rev3 (builtin pump and res) vs XSPC RX360?


And about HWLabs GTX 360... I'm confused by the different dissipated watts numbers on skinnee and martins sites:
http://martinsliquidlab.org/2012/04/09/hardware-labs-gt... - 1.5GPM/1400RPM = 205W
http://skinneelabs.com/hwlabs-gtx-360-radiator-review-d... - 1.6GPM/1200RPM = 466W (more than 2x!!)
What am I missing? Why the numbers are so different?

wrt TDP calculations:
i7 3930K (stock) = 130W
i7 3930k (OC/4.5G) = 250W
GTX 680 (stock) = 190W
GTX 680 (OC, didn't find numbers) = 220W (approx)
I don't think I ever need both at 100%, so 2 GTX680 + OC CPU roughly translates to 550-600W. Does that sound right?

wrt Reservoirs: I didn't see decent comparisons for reservoirs.. I saw a post on the Martin's website, when he got good results even with MCP35X Reservoir top. Are there any well-known good res's out there?

wrt CPU blocks: Looks like XSPC RayStorm is comparable to Apogee XD, but all reviews say Apogee has much nicer and better mounting. It also has 2x restriction comparing to XSPC RayStorm :/  How bad is it?

And speaking about pressure. According to my (completely amateur) calculations:
MCP35X + Apogee HD + 2 * MCW82 GPU blocks + 2 MCR320-QP = ~1.2GPM
MCP35X + XSPC RayStorm + 2 * MCW82 GPU blocks + 2 HWLabs GTX 360 = ~1.2GPM
Does that sound right? And how good is that flow?
m
0
l
a b à CPUs
a c 78 K Overclocking
April 19, 2012 1:05:45 PM

low restriction, good mounting and material choice determine how well a block will perform. since the HD has a good contact patch, better than all the blocks compared it trades blows with other blocks that have reduced contact. so you understand now, why the HD is a good perform inspite of having higher restriction?

you could also look at the DT 5Noz...but is pricier

rads, the heat source is what is different from these tests, have you also noticed the ambient temps of both test procedures?

the other memebers can point out whats cooking with the two dif tests results+blocks :) 
m
0
l
!