|
|
#1 |
|
...
Join Date: Feb 2008
Posts: 3,026
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Air vs Water - Price/Performance Ratio
Well not sure I need to write anything here, the title says it all.
I have to buy a new cooler for my graphics card (silly me) so I am either going to wait a week and get water cooling, or just get a £20 thermalright cooler for the card. So really if I water cooled my cpu, maybe chipset and graphics card, how much better are the temps likely to be than air cooling. Lets say I use a 120x2 or 120x3 rad, that means including my 2 x 120s on my case at the moment I would have 4 or 5 fans. Now my freezer 7 pro is incredably quiet, so I wouldnt count that as a factor in noise, so lets just say I got a good semi passive cooler like the arctic cooling or thermalright ones with a slow spinning fan on my graphics card, my system would be pretty quiet. Now I fear that if I fork out x amount on water cooling, it may be just as loud, if not louder due to the fact I am going to have to mount my rad externally, and the extra sound of a pump (no matter how quiet it is, its still extra noise). So really how much better are the temps on average, I realise the question is flawed in so many ways, but I am sure a few water coolers could give some before and after examples. Anyway considering my first line I think I managed to write quite a bit. Cheers chaps |
|
|
|
|
|
#2 |
|
Waste of space
Join Date: Mar 2007
Location: Pittsburgh , USA
Posts: 41
![]() |
The cost/performance ratio differs greatly depending on your system.
If you have decent air cooling and a case that has good flow Your not going to see an increase from water cooling (compared to the price difference) But if you have a cluttered case adding water with an external rad will show a big difference in price/performance. |
|
|
|
|
|
#3 |
|
Super Moderator
Moderator
Join Date: Jan 2003
Location: Lancs, UK
Posts: 22,157
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
As above - the performance of WC varies massively, depending on exactly how you set it up: it's both its greatest strength and weakness.
To me a completely WC'd graphics card is pointless, because you need a fan blowing over it to maintain a decent, even temperature (or even-ish anyway). Full-cover blocks work quite nicely, but are a ballache to fit and very expensive to buy, because you have to change them for (nearly) every graphics upgrade. I'd say you should spend a decent chunk of cash on a good aircooler, make sure it's as quiet as you want it, and you're laughing.
|
|
|
|
|
|
#4 |
|
...
Join Date: Feb 2008
Posts: 3,026
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Well I would say my case is pretty neat, as of a few days ago.
According to RealTemp, which I think is the most reliable way of measuring my E8200 according to many sources, my cpu overclocked to 3.4ghz at idle is 36/37c, and usually goes no higher than 45c. My graphics card cooler is bust atm so I cant comment on the temps, lets just say to be using it now I cant go near a 3d app and I have underclocked it, and its doing fine. |
|
|
|
|
|
#5 |
|
...
Join Date: Feb 2008
Posts: 3,026
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Cheers krikkit, I was already responding to the first post before you nipped in there.
Yeah I was looking at how pricey full cover blocks are, and plus the EK ones (which seem to be the only ones available for my card) dont fit my PowerColor card because of the different layout, EK told me it would fit but I would have to remove the power regs cooler part, and then use the standard one, but this would surely require some degree of airflow, thus slightly defeating the point. So really I think this time round water cooling is not for me, thats 2 builds I have put off water cooling for. I think I just like air way too much. Plus I really like the look of a lot of the graphics card air coolers. Maybe I just wanted someone to tell me to air cool so I could justify the decision I had already come to in my head. |
|
|
|
|
|
#6 |
|
Resident EI
Join Date: Mar 2006
Location: Almost Heaven (WV)
Posts: 870
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Here's the low-down from an avid watercooler.
Benefit/cost ratio of watercooling vs good aircooling is dependent on several factors; most importantly the components to be cooled and if/how much you plan on overclocking. Low-end to mid-range GFX cards won't see as much of a benefit from water (compared to good heatpipe coolers) as high end hardware will. Also, if you plan on running the system at stock you'll only see lower temps but not really any performance gains; however, for substantial overclocking, watercooling can maintain significantly lower temps that aircoolers ever can and thus allow you to overclock further without running into thermal throttling issues. For watercooling the graphics cards, the core-only blocks + RAMsinks are most versatile, but generally occupy 2 or more PCI slots (and the RAM really doesn't get overly hot and doesn't necessarily need active cooling); Full-cover blocks offer an elegant solution and invariable only use 1 PCI slot, but can be quite pricey. The temps on my 8800GTXs actually dropped over 10C with full-cover blocks compared to the stock cooler. For an example on how watercooling can help overclocks: I had a P4 540J 3.2GHz that I got to 3.76GHz on stock cooling with full load temps around 60C, after putting it under water I got it up to 3.92GHz at 40C; I could have pushed that CPU even further, but the chipset temps were getting a bit high. The one word of warning I have for anyone considering water cooling is to avoid cheap components from companies such as ThermalTake and Koolance; they generally only perform on par with good heatpipe coolers, some parts are more prone to failure, and they often use Aluminum in their blocks and/or radiators which leads to corrosion issues (between the copper bases of the blocks and the aluminum). The only manufacturers I'll consider for H2O components are: DangerDen, EK-Waterblocks, Swiftech, D-Tek, Thermochill, Hardware Labs, and LIANG. The pumps are often extremely quiet; I'm running a LIANG D5 (aka Swiftech MCP-655) and it's near silent, it's actually significantly quieter than the fans and HDDs. |
|
|
|
![]() |
| Tags |
| cooling |
| Thread Tools | |
|
|