iXBT Labs - Computer Hardware In Detail

Platform

Video

Multimedia

Mobile

Other

Integrated assessment of consumer qualities of cooling systems



Detailed descriptions of cooling systems, thorough examinations of their technological peculiarities and operating abilities, multiple results of various tests are undoubtedly useful. But it happens quite often that a user can't choose a cooler even after he has read hundreds of reviews and looked through a heap of pictures. 

Well, the common disadvantage of most reviews of cooling systems for computers is lack of a single integrated assessment of functionality and quality of such systems which would allow grouping all technical parameters and making an informative and comprehensive comparison characteristic of coolers. Of course, there are some sites over the Net offering charts and ratings, but it is vague how such ratings are formed and what objective (subjective) parameters are accounted for. When devices of doubtful quality win and worthy ones lag behind it is clear that the given approach is inapplicable. 

We decided not to give up and elaborated our own technique of integrated assessment of consumer qualities of cooling systems. We do not consider it ideal but it reflects an actual performance of a certain cooler to some degree of objectivity. 

So, here is the idea of the technique. 

Any cooling system has 6 main estimated characteristics: 

  1. Thermal efficiency (thermal resistance θja
  2. Noise level
  3. Technical quality of active components (fans, rotary pumps in liquid cooling systems etc.) - estimation of technical parameters of such devices, their functionality and reliability
  4. Technical quality of a mounting mechanism - estimation of handy installation, technological effectiveness of a fastener, its reliability and conformity to the norms
  5. Technical quality of passive components (heatsinks, heat exchangers, pipe lines etc.) - estimation of technical parameters, technological perfection, reliability
  6. Price attractiveness (price effectiveness) - estimation of a cost of a cooling system and how this parameter relates to current economic realities
There are two bonus categories in addition to the above characteristics: 
  1. Technological effectiveness of a system (estimation of thermocontrol scheme, fan failure alert, adapters/adjusters of power supply, protective mechanisms etc.) 
  2. Accessories (thermal interfaces, power adapters, guides and descriptions, additional gadgets and devices) 
Each of the characteristics gets its own index which is calculated according to the established procedure. 

Thermal efficiency

Eth = (WC*RFth)/TR (scores),  

Eth - thermal efficiency index, WC - weighting coefficient, RFth - thermal reference factor (reference thermal resistance θja), TR - thermal resistance θja of a cooling system in question. 

The reference thermal resistance is 0.30°Ñ/W for Socket 478 systems, 0.25°Ñ/W for Socket A systems (the best values regarding the modern technological capabilities of mass production). The weighting coefficient equals 60. 

Noise level

N = (WC*RFn)/NL (scores),  

N - noise index, WC - weighting coefficient, RFn - noise reference factor (reference noise level), NL - noise level produced by a cooling system. 

The reference noise level is 25 dBA (typical background noise level indoors at night). The weighting coefficient is 60.

Technical quality of active components (TQ AC) and technical quality of a mounting mechanism (TQ MM)

"Unsatisfactory" - 0 scores
"Satisfactory" - 5 scores
"Good" - 10 scores
"Nothing to find fault with" -15 scores

Any intermediate marks are also possible. For example, 14 scores can be interpreted as "the fan quality is excellent, but the power cables have too dull colors (the fastener is wonderful but its mirror surface doesn't reflect my face correctly :)", and 1 score can mean that "the fan quality is awful but it is still able to work a couple of works making a lot of noise". 

Technical quality of passive components (TQ PC)

"Unsatisfactory" - 0 scores
"Satisfactory" - 4 scores
"Good" - 7 scores
"Nothing to find fault with" -10 scores

Any intermediate marks are also possible.

Price attractiveness

Ep = (WC*RFp)/P (scores),  

Ep - price effectiveness index, WC - weighting coefficient, RFp - price reference factor (reference cost), P - average retail price of a cooling system. 

The reference cost is 7 euros for systems of forced air cooling (most coolers), 12 euros for improved coolers (equipped with hi-tech heatsinks), 35 euros for liquid cooling systems. The weighting coefficient is 60.

Bonuses

Correctly working thermocontrol - 15 scores
Fan failure alert - 5 scores
Preinstalled thermal interface - 5 scores
Power adapter (adjustable) - 5 scores
Thermal grease supplied - 3 scores
Others - 1 score 

All indices and bonuses are calculated after examination of each cooling system and brought into its "consumer map". The sum of all indices and bonuses gives "an integrated assessment of consumer qualities of a cooling system" (IACQ CS). If a price attractiveness is not taken into account (for example, for estimation of pure functionality of a cooler), "an integrated assessment of functionality of a cooling system" is calculated (IAF CS). In the end all obtained IACQ CS and IAF CS are taken to make the Table of Ranks - consumer and technical ratings of cooling systems. 

Well, I hope that the practical application of this new technique (the touchstone is the review Coolers for Socket 478, Spring/Summer 2002) will help you to solve this complicated problem of choosing a decent cooling system and a solution which would meet the toughest requirements of quality and functionality. Comments and suggestions are welcome!  
 
 

Vitaly Krinitsin (vit@ixbt.com)

Write a comment below. No registration needed!


Article navigation:

blog comments powered by Disqus

  Most Popular Reviews More    RSS  

Creative Sound Blaster X-Fi Surround 5.1

An external X-Fi solution in tests.
September 9, 2008 · Sound Cards

ASUS M5A97 EVO, ASUS M5A99X EVO Motherboards

Mainstream Socket AM3+ boards from the new M5 series.
November 1, 2011 · Motherboards

71 CPU Roundup

Summing up the year 2011.
January 25, 2012 · General Platform

ASRock P67 Pro3 (B3) Motherboard

A mid-end model with USB 3.0, eSATA 6Gbps and UEFI.
March 29, 2011 · Motherboards

Gigabyte GA-890FXA-UD7 Motherboard

AMD 890FX chipset in tests.
May 17, 2010 · Motherboards
  Latest Reviews More    RSS  

i3DSpeed, April 2012

Retested all graphics cards with NVIDIA Drivers 301.24 and AMD CATALYST 12.4. Added test results of the reference and overclocked AMD Radeon HD 7850, NVIDIA GeForce GTX 690, NVIDIA GeForce GTX 680 SLI, AMD Radeon HD 7970 CrossFireX, AMD Radeon HD 7770/78
May 05, 2012 · 3Digests

i3DSpeed, March 2012

Retested all graphics cards with NVIDIA Drivers 295.73 and AMD CATALYST 12.3. Added test results of the reference and overclocked AMD Radeon HD 7870 and NVIDIA GeForce GTX 680.
Apr 05, 2012 · 3Digests

i3DSpeed, February 2012

Retested all graphics cards with NVIDIA Drivers 295.52 and AMD CATALYST 12.1, added test results of AMD Radeon HD 7970/7950/7770/7750/6930.
Mar 05, 2012 · 3Digests

Palit GeForce GTX 560 Ti Twin Light Turbo 1024MB GDDR5, KFA2 GeForce GTX 560 Ti LTD OC 1024MB V2.0 Graphics Cards

A couple of interesting custom cards, one heavily overclocked.
Mar 01, 2012 · Video cards: NVIDIA GPUs

i3DSpeed, January 2012

Retested all graphics cards with NVIDIA Drivers 295.52 and AMD CATALYST 12.1, added AMD Radeon HD 7950 test results.
Feb 14, 2012 · 3Digests
  Latest News More    RSS  
  Useful Links Get listed  

Wholesale Computers & Networking

Get great Dell Coupons at CouponSnapshot.com

Saving more with great Lenovo coupon codes

Cut your budget with Coupon codes

Great HP vouchers

Platform  ·  Video  ·  Multimedia  ·  Mobile  ·  Other  ||  Feedback  ·  About us & Privacy policy  ·  Twitter  ·  Facebook


20

Copyright © Byrds Research & Publishing, Ltd., 1997–2011. All rights reserved.