ASUS GeForce GTX 960 Strix Video Card Review


<< PREVIOUS            NEXT >>

Video Card Testing Methodology

Due to Microsoft DirectX-11 (DX11) being native to the Microsot Windows 7 Operating system, it has been our primary choice for our test bench for quite some time. The DX11 API is also available as a Microsoft Update for the older Windows Vista O/S, and it is native on the newer Windows 8 and Windows 8.1 Operating Systems.

By Benchmark Reviews high standards, we always try to take a bigger sample and do simple statistical procedures to obtain the most accurate results that represent how each item performs in a certain area. Each benchmark test starts with a “cache” run, followed by five recorded test runs. From those five scores, the two outliers are eliminated and the average from the remaining three is calculated and displayed as a score in the following pages.

Our combination of synthetic, video game, and compute benchmark tests has been chosen in this article to correctly represent relative performance among video cards. These benchmark results are not intended to represent real-world graphics performance, as this can change based on different system configurations, and the perception of individuals that play the video game.


Intel Z87 Test System


CPU Intel Core i5 4670k @ 4.0 GHz
CPU COOLER Raijintek Triton 
MEMORY 8GB G. Skill Ripjaws 2000 MHz ram
STORAGE DEVICES WD Black 2 TB DriveSamsung MZ7TD120HAFV 120 GB 
MONITOR QNIX 2710LED 2560×1440

Video Cards Being Tested

Driver Memory Amount Memory Interface Memory Clock (MHz) Boost Clock (MHz) Base Clock (MHz) GPU Cores
Gigabyte GTX 980 Forceware 347.09 4096 MB GDDR5 256-bit 7010 1216 1126 2048
Evga GTX 970 SC Forceware 347.09 4096 MB GDDR5 256-bit 7010 1367 1165 1664
ASUS GTX 960 Strix Forceware 347.25 2048 MB GDDR5 128-bit 7200 1291 1228 1024
XFX DD R9 290x Catalyst 14.12 4096 MB GDDR5 512-bit 5000 N/A 1000 2816
XFX DD R9 280x Catalyst 14.12 3072 MB GDDR5 384-bit 6000 N/A 850 2048
GTX 770
Forceware 347.09 2048 MB GDDR5 256-bit 7010 1137 1085 1536
Evga GTX 760 Forceware 347.09 2048 MB GDDR5 256-bit 6008 1033 980 1152
Evga GTX 750 TI Forceware 347.09 2048 MB GDDR5 128-bit 5400 1085 1020 640
Evga GTX 660 TI Forceware 347.09 2048 MB GDDR5 192-bit 6008 980 915 1344


<< PREVIOUS            NEXT >>


Skip to comment form

  1. Bob

    Was disappointed to see no noise measurements. Have read of numerous complaints about coil wine and fan noise on many of the Nvidia 700 and 900 series cards, so when you mentioned that Asus was claiming 3 times lower noise for the Strix, I was anxious to find out exactly what they meant.

    Also, the fact that the card has no fans running at all until it reaches 55 degrees Celsius had me wondering what it would be like to have a totally silent card that suddenly activated it’s fans at a certain temp. Would that activation be jarring or annoying or would it be a gradual ramping up, and how loud would it be?

    I’m still running a GTX 560 ti and am ready for an upgrade when the right card comes along, so am really curious about the noise levels of this one. Although I agree with your suspicion about the odd price point, and that there is usually something in the $250.00 – $275.00 range that is missing in the 900 series.

    1. Caring1

      The lower noise can be claimed due to the fans not running during idle, it is a marketing trick as the fans will not be that much quieter usual when actually running.

      1. Bob

        I suspected that as well, which made it even more unusual that Julian reported no noise measurements.

    2. Caring1

      I was a bit disappointed not to see an R9 270 included in the test, as in my opinion that is it’s direct competitor, judging by size, price and performance.

    3. Julian Duque

      Hello Bob. We did not record noise measurements in this article as we will soon be doing a complete noise roundup test soon with the appropriate tools. Just to recap something I didn’t mention, the fans never ran at 100% during all of our tests.

      1. Bob

        Will be waiting anxiously for the “noise roundup test”, but it should be part of every video card test.

        1. Julian Duque

          We understand, but we are in the process of obtaining better equipment to test for noise using quantified measurements instead of the usual Max noise score that we used in past reviews. We did not want to delay this review past today, and that is the reason why there are no noise tests.

        2. Olin Coles

          Yes, everything measurable should always be part of every review… unless you’re only given three days (evenings) to complete a lengthy project you offer to the public completely free of charge.

          While I’m thinking about it, you know what should be in every comment below a review? Some form of gratitude along with whatever question you have.

          1. Roman from USSR

            Indeed, I totally agree with Olin. People used to only complain/ inquire in the comments.
            Thank you for your reviews guys, they are helpful.

  2. Tom J

    Thanks for having this out so soon! You guys beat Anandtech 😉 (Along with almost everyone else!) – I know you probably didn’t have time, but any word on how this thing overclocks (and how it performs when overclocked)? With the other Maxwell cards able to hit 1300/1400 boost clocks, I’d be curious to see if the 960 would as well. Could change the equation a bit, although overclocking definitely varies per card. Seems like it’d be easier to jump to an overclocked R9 290 otherwise and get 970ish performance for a bit more (although, if you want to talk about noise…) Appreciate the observations on the new Maxwell, thanks Julian!

  3. edy

    “”At the center of the PCB we find the star of this review, the GM 204″”

    Really? Isn’t the GM 206?

    1. Julian Duque

      Thanks for bringing this up, I have now corrected the article. It is in fact the GM 206, as the GM 204 is found on both the GTX 970 and 980.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>