Additional Fan Tests

Initial test results showed the fan of the Vindicator was heavily slanted to silence. The stock fan provided average to below average results in our cooler overclocking tests. To determine whether the overclocking limitations were a result of the cooler design or the fan selection, additional cooling tests were run with a remarkable new fan we just got in for review.



The SilenX IXTREMA 120 claims to provide the world's best noise-to-airflow ratio. The specifications certainly support that claim with an exceptional 14 dB-A noise level coupled with 72 CFM airflow. Perhaps even more remarkable is the rated watts of 1.92 which should be safe on almost any motherboard fan header.

We extend our sincere thanks to Frozen CPU for providing the SilenX IXTREMA 120 for testing. We are preparing for a roundup of 120 fans and Frozen CPU has provided an assortment of tantalizing fans that we think you will enjoy seeing in the upcoming 120 fan roundup.

The claimed low noise and high output of the SilenX fan certainly sounds attractive, so the IXTREMA was used to see if higher fan output would improve the cooling and overclocking of the OCZ Vindicator. Results with a single stock OCZ fan and a single SilenX IXTREMA 120 are reported in all performance charts and graphs.

Push-Pull Testing

Since a push-pull fan configuration made a significant improvement in our review of the Scythe Infinity cooler, push-pull was briefly tested using two stock OCZ Vindicator fans. Overclocking performance was also tested briefly using two SilenX IXTREMA 120 fans.

Push-Pull results were mixed and not directly comparable to the findings on the Infinity. A stock single OCZ Vindicator fan topped out at an overclock of 3.80Ghz. Using two of the 40 CFM fans allowed us to reach a top stable overclock of 3.87GHz - better than one fan, but still below the top heatpipe tower coolers and the push-pull Infinity.

However, a single 72 CFM SilenX allowed a top stable overclock of 3.90GHz - matching the top results with the Thermalright Ultra 120/SFLEX, Tuniq Tower 120, Cooler Master Hyper 6+, and Scythe Infinity (with two push/pull 46.5 CFM fans). This certainly points to a better fan providing better results with the OCZ Vindicator. The soon-to-be-released Thermalright Ultra 120 Extreme remains alone at the top of our performance charts with an overclock of 3.96GHz combined with temperatures among the lowest tested at each overclock.

With these results with a single SilenX we were excited to test the OCZ Vindicator with two SilenX fans. To our surprise two SilenX 72 CFM fans in a push-pull configuration provided no better overclocking than a single 72 CFM SilenX. The conclusion is that increasing airflow in the Vindicator definitely improves performance - moving it from average to a top competitor. However, increasing airflow, even with high output low-noise fans, is not without limits. Eventually you reach the point where increasing airflow provides no further improvements in cooling or overclocking. This optimum point is likely to vary with each cooler and each type of design. It is a subject we will explore further in our 120mm fan roundup.

Since results were the same with one or two SilenX fans, we report results with one SilenX fan and the stock OCZ Vindicator fan. This is also consistent with the fact that two OCZ fans in push-pull improved cooling, but not to the extent of the improvement with one SilenX IXTREMA fan.

CPU Cooling Test Configuration Cooling at Stock Speed
Comments Locked

41 Comments

View All Comments

  • Wesley Fink - Thursday, April 5, 2007 - link

    You make a good point, but I remain concerned about introducing additional noise and cooling variables into the test platform with a variable fan speed PS. In flat out gaming every PS I have seen is running at load conditions. However, your suggestions are certainly "real world" and we will definitley evaluate the impact of one of the "silent idle" power supplies on our total cooler testing. Our goal is to find ways to lower the ambient noise and sytem noise floor in realistic ways to provide a wider db test range
  • yacoub - Thursday, April 5, 2007 - link

    Yeah exactly. My PC DOESN'T sit at Load settings for its fans. See I actually have them all dynamically controlled by temp and they only spin up when the machine is actually processing heavily enough to warm things up to a level that needs the fans to spin up a bit more to cool effectively.

    I'm starting to think it's more a matter that most folks don't care enough to quiet their PCs down so they don't realize how much nicer it is to have a quiet PC. This would be why they think it's normal to have a PSU making 40db of noise and thus fans can be 36db of noise and they think that's ok.

    My PC (and it's not exactly designed to be totally silent) is inaudible except when I'm gaming or doing heavy work in a program in windows that requires the cpu to load up a lot. I can surf the Internet, type a Word doc, listen to music, or watch a DVD, and it's inaudible unless you put your ear within a foot or two of the case.

    (P150 Antec case, Antec True-Power2.0 450w PSU (recently replaced by a Corsair 520w that has a fan that is a tiny bit more audible), SpeedFan, ATiTool, and good configurations are to thank for that. It's even using the retail cooler on my A64 Venice 3200+ and it's overclocked to 2.4GHz and it's still silent because I have the fan on it throttle down to 40% when it's under 40C, which it is most of the time.)
  • punko - Thursday, April 5, 2007 - link

    For me, my wife watches television in the same room that I have the computer. When I'm gaming I use either very low sound or headphones.

    Noise is an issue, whether at idle, low load, or heavy load.

    If anything, I tend to have the CPU and GPU maxed out when I'm on the computer, so the "real world" testing of a complete system underload is the most important sound rating. Idle? don't care, that means I'm not in the room.

    What I would love to see is a general guide for those of us planning to build a new rig, and get an idea of what is available in terms of major components for building a quiet (not silent) PC. I am more than well aware that there are specialist sites that do component by component reviews, but I am more interested in the "real world" knowledge of people that assemble and use hundreds of components in various configurations.

    Articles like this one help, because the focus the attention on a particular component as part of an overall system.
  • yacoub - Thursday, April 5, 2007 - link

    oh and by "only spin up" i mean "spin up from the level I have them throttled down to" not "spin up from stopped". You never want to throttle fans down to where they stop while the system is on, because there's always the danger they won't spin back up.
  • asdfqwertyuiop - Thursday, April 5, 2007 - link

    This article is poor. The SilenX claims are just false, and parroting their marketing BS without adequate testing is unsatisifactory. There are about a dozen fans, including the stock Intel one, all on the same db rating. So in other words, testing was insufficient to establish whether or not the fans are quieter than even the bog-standard Intel one.

    Accepting SilenX's BS that their fan achieves dramatically more airflow with less noise than anything else available shows excessive credulity.

    Aside, what's with tech firms and the product names that sound like they are aimed at 13-year olds? Vindicator??? This is a heatsink, not an assault weapon.
  • tomoyo - Thursday, April 5, 2007 - link

    As stated by quotes from a silentpcreview reader here - http://www.silentpcreview.com/forums/viewtopic.php...">http://www.silentpcreview.com/forums/viewtopic.php...

    There's so many inaccurate and incorrect commentary in this review that it's utterly amazing it was ever posted. It's obvious the heatsink was designed by scythe and relabeled by OCZ, that's nothing terrible to say since tons of brand names are made by different companies. I'm MOST disgusted by the blatant love of Silenx fans when their specifications on their fans are 100% WRONG. It is physically impossible for their fans to be as quiet as they say, which means they're purely and utterly liars. A company that lies should not be praised like anandtech is doing. I'm surprised they are not using a quiet nmb-panaflo, nexus, or other good fan out there.
    And the OCZ powerstream is certainly not one of the quieter power supplies out there. A better comparison psu would be something like a Seasonic or Corsair.
    There's plenty of other issues with the writing quality here, but I think these are the most blatant.
  • amdlive - Thursday, April 5, 2007 - link

    quote:

    I'm MOST disgusted by the blatant love of Silenx fans when their specifications on their fans are 100% WRONG.


    What is the true specifications of the fan utilized in this article? If no the specs listed are 100% wrong, then provide the real world results you have. We are waiting and by the way, post a link to your last review.
  • amdlive - Thursday, April 5, 2007 - link


    quote:

    I'm MOST disgusted by the blatant love of Silenx fans when their specifications on their fans are 100% WRONG.



    What is the true specifications of the fan utilized in this article? If you know the specs listed are 100% wrong, then provide the real world results you have. We are waiting and by the way, post a link to your last review.
  • goinginstyle - Thursday, April 5, 2007 - link

    Silent PC Reviews are a total joke. They in no way simulate conditions inside, near, or even remotely in plain sight of a PC. Most of the members over there are elitist wannabes who drink the Chin Kool-Aid and then march off like ants to spread the gospel of lies that are the SPCR test results. Most of their articles are contradictory once you really read them and the computer equipment they use is several generations old. They are not even using equipment that most of us have bought in the last two years and to even consider a P4 and 6800GT as components that stress a power supply is just laughable. The site is one big cluster "F" with most of the recommendations going to their sponsor list. I could not believe some of the asinine comments posted in that topic. Hopefully they wake up and smell the $h!t that is spewing from the writers over there.

    My issues with the cooling reviews here are the noise levels in the room. They really need to be around 26 to 30dBA to properly test most components in your "real world" test routines. There are a lot of people who probably have rooms at home in that noise range along with components whose base noise levels fall in the same category. I would also like to see how well the units fit on the top three or four selling motherboards along with some power supply reviews.
  • Wesley Fink - Thursday, April 5, 2007 - link

    According to a Noise article at SPCR 26 to 30db is the noise level of a "quiet bedroom at night". 35 to 45 db - our ambient noise is 36db - is the noise level of a "Quiet office or library; Typical PC Subdued". The complete chart is in comments below. We hope to lower noise a bit in a reworked test bed but we are not confident we can get below 30db consisitently.

Log in

Don't have an account? Sign up now