Source: Tom's Hardware UK – Keywords: CPU-Coolers, Phenom, Core2
Categories: Hardware
The Conclusion - Failure Rate of 45% - Thermalright and MSI/Watercool Recommended
It really is remarkable how much junk is sold in hardware stores these days and how badly customers are sometimes deceived. Any specifications quoted on the boxes should be taken with more than just a pinch of salt as they very rarely coincided with our findings in the lab. After all, you can’t tell a cooler’s performance or noise level simply by unpacking it and looking at it closely. It is therefore highly unadvisable to buy a cooler without having read at least one independent review of it if you care about getting good results. The sobering balance of first part of our test covering more than 80 CPU coolers is this – 45% of the products tested here either fail our tests or simply aren’t worth their price.
The width, length or number of heatpipes alone is not a good indication of a product’s cooling performance. Just as important, if not more so, is the composition of the gas inside the heatpipes and with what type of cooling configuration the heat is dissipated into the surrounding space.
Thermalright offers a very good solution with the IFX-14, a model that provides very good cooling performance. Even with all four cores of our quad-core processor under full load, core temperature reached only 62°C. Paired with the Scythe SY1225SL12M fan, the cooler is all but silent in operation. As a result, it receives our recommendation.
We were also especially impressed with the water cooling solution presented by MSI and Watercool. Installing the HydroGen/HT Fusion Dual is extremely easy, and the cooler is ready for use as soon as the reservoir is filled. There are no problems getting the system started, as water flows through the system instantly. Also, the cooling unit can either be placed on top of your computer case or on the floor. Additionally, the cooler can be monitored through a standard fan header on the motherboard – not something that is to be taken for granted in an external water cooling solution. The pump’s performance can be adjusted to your needs as well. Its ease of use combined with the good cooling performance win it our second recommendation.
Due to the immense number of products in this test field, we have split this roundup into three parts.
We were disappointed by manufacturers Scythe, Silverstone, Zerotherm, and Zaward. Their products failed our test due to insufficient cooling performance, inferior build quality or an ill-conceived installation procedure.
- Previous page Weight and Fan Speed
- Ultimate Budget Overclocking Box - A 3.5 GHz Core 2 System with a...
- Finding The World's Best Hardware Prices: Shop Globally
- Comparing Water Coolers: We Follow Your Lead
- Keeping Your CPU Going If Your Cooler Fails
- Cooler Master CM690 : The happy medium?
- How Cool Are Thermalright's Graphics Card Coolers?
- Extreme FSB: Taking the E6750 Beyond 4 GHz
- Radical CPU Coolers from CoolIT
- Computex 2007 Day 1: Coolers and Power Supplies
- A Beginner's Guide For WaterCooling Your PC
I think people reading this need to bare in mind 2 things:
1) the test is using a extremely high power processor, if you were to test these coolers on a dual core CPU you may see very different trends with regards to noise/performance. The ninja is famous for quiet cooling at lower settings, passively there are few better heat sinks than it.
2) the test is not been done in an enclosure, in some cases the very high level or very low level of airflow could make a drastic difference to the performance of some of these coolers.
The Noctua can be installed as above (airflow vertical) or twisted 90 degrees (air horizontal). Most people I expect would install it for horizontal air flow, so the air would be vented straight out the back of the case. Would this have an effect on the efficiency of the heatpipes??
With a case that has a good airflow like the Antec P18* the Ninja works fairly well, how would the case´s airflow affect the cooling performance?
My p180b with the Ninja plus rev B and a Noctua NF-S12-1200 works well with the Q6600, yesterday I did a Prime session, after 1h max Tjunction was 61 Core 1, core 3 was the coolest at 57.
Can you tell me why you did not test the Scythe Ninja Plus Rev. B with 4 fans as anyone who would be using a Core 2 Extreme cpu and hoping to overclock would instantly purchase at least an extra fan for the Ninja. I myself have the scythe Ninja Plus Rev. b and I brought it for 2 main reasons, it can be used fanless (this was never stated in this article) and you can add up 4 120mm fans to increase cooling effectiveness (again this was not stated in the article and its a big plus to anyone whose looking to buy one).
Can you clearly explain why you choose not to do the test in a enclosure. Surely you didn't take into account the effect intake and exhaust fans, they are the key things to helping the heatpipes working at there best. The cool air coming from the intake fan is what aids the heatpipes and the exhaust remove the hot air, again aiding the heatpipes to work.
When you try them in open air you lose a lot of the power of the heatpipe as there isn't as much of a contrast between the cold and hot air.
Guys pull me up if I'm wrong, but if I'm not you really should redo this test in a enclosure.
See above it does say "Part I", they might test the Ninja later..
I would like to see the ZALMAN 9500A 92mm 2 Ball CPU Cooler in the test at some point. I have found this to be a good cooler and many people I know like it as well.
I really can't understand why these reviews have been carried out with the fans trying to blow hot air back downwards through the cooling fins!!
I mean look at the pictures on the Scythe Ninja page of this review for example - they just scream out "WRONG!!!!!".
Even an 8 year old will tell you that hot air rises!! The ideal placement for the fan is therefore either on the side for horizontal airflow through the fins or underneath so the fan actually helps the already rising hot air to rise even quicker. I reckon you could easily drop most of the recorded temperatures by up to 10 Degrees or maybe even more by mounting the fan(s) anywhere but on top blowing down.
If you really had no choice but to mount on top because of interference from other motherboard components (in which case the cooler should lose points), then at least have the fan blow the hot air away from the heatsink fins and not back through them!!
Kulwant this is not always the case. There are many desktop cases that have a grill in the case over the CPU socket allowing fans to suck cold air from outside the case and blow it down on to the hot heatsink - thereby also cooling the voltage regulators. If you swap the fan round to blow upwards then it would be sucking warm air from inside the case to draw over the heatsink fins which would reduce cooling performance. Also it would be working against the flow in tower cases which have fans on the side blowing air in.
This review does lack the Zalman 9500 series of fans as Therlian mentioned (including the Fatal1ty versions) but I think a nod to them was placing 2 on the motherboard in the Skulltrail articles - clearly one of the best air coolers around
David_T, I appreciate some designs blow cold air brought in through a side panel and through the CPU heatsink - but you'll usually find that the flow of air is still horizontal and not top->down. Especially in the normal vertical orientation of the motherboard (i.e tower upright - not lying down). I'm not convinced that such designs necessarily blow cold air through the voltage regulators as once the air's been through the CPU heatsink - it's not exactly cold anymore. And I've yet to see a design that had a specific stream of cold air just for the voltage regulators.
The heatpipes favoured by a lot of the P35 chipset boards seem to be about simply getting some airflow through the voltage regulator heatsinks not necessarily cold airflow.
OK, I admit my comments don't necessarily apply to all heatsinks tested here - but the point is that in some cases I believe the tests are flawed and don't reflect the true potential of some of the solutions (e.g. the Scythe Ninja and possible some others too) simply because of the sub-optimal placement of the fans.
Hey people I think we have to reconsider our opinions about Zerotherm. Watch this http://www.pchs.it/forums/cooling/ [...] btf92.html