«

»

Crucial Ballistix Elite DDR4-2666 Memory Review

PAGE INDEX

<< PREVIOUS            NEXT >>

Testing and Results

Testing Methodology

As mentioned before, I’ll be doing a head-to-head comparison between the Crucial Ballistix Elite DDR4-2666MHz, and the Corsair Vengeance LPX 2400MHz Memory kit. I used their XMP profiles to set their timings, which my motherboard had no issues setting and were perfectly stable.

For comparison sake, I’ll list the timings of each RAM kit:

Crucial Ballistix Elite: 16-17-17-36 CR-1 2666MHz

Corsair Vengeance LPX: 14-16-16-31 CR-2 2400MHz

Test System

  • Motherboard: GIGABYTE GA-Z170X-Gaming 6 (rev. 1.0)
  • System Memory: Crucial Ballistix Elite DDR4 16GB 2666MHz / Corsair Vengeance LPX 16GB 2400MHz
  • Processor: Intel Skylake 6700K @ 4.2GHz Stock settings
  • Audio: On-board Realtek ALC-1150 codec
  • Video: EVGA 980 Ti Classified Edition
  • Disk Drive 1: Samsung 840 Evo 120GB SSD SATA 3
  • Disk Drive 2: Western Digital Black 2TB HDD SATA 3
  • Optical Drive: Asus DVD/Burner
  • Enclosure: Cooler Master HAF 922
  • PSU: Corsair HX850 850 Watt
  • Monitor: Samsung 32″ LCD
  • Operating System: Windows 10 64-bit Home

Test Software

  • Aida 64 Engineer 5.75.3900
  • SiSoft Sandra Lite 2016.03.22.20
  • Passmark Performance Test 9.0
  • CPU-z 1.76.0
  • Handbrake 0.15.5.0 64 bit
  • 3D Mark 11 Profesional Edition
  • Heaven Benchmark 4.0
  • Valley Benchmark 1.0

Results

AIDA 64 Engineer 5.75 Memory Test

Aida 64 is one of the main staples of synthetic tests for finding maximum theoretic memory bandwidth. As with all synthetic benchmarks, this isn’t an accurate means to measure real-world performance. But, for the purposes of comparison in this article it will suffice.

Crucial-Ballistix-Elite-Aida-64-Engineer

The higher bandwidth Crucial Ballistix Elite wins out in all 3 categories. Crucial Ballistix Elite manages an 11.82% increase in read, 10.71% in write, and 9.51% in copy. Synthetic maximum bandwidth should be taken with a grain of salt, as it usually bears little real-world results.

SiSoft Sandra Lite 2016.03.22.20

SiSoft SANDRA (System Analyzer, Diagnostic and Reporting Assistant) compromises a whole suite of benchmarking tools to analyze the performance of various hardware components. That is outside the scope of this article and we’ll be focusing on the memory tests specifically.

Crucial-Ballistix-Elite-SiSoft-Sandra-Lite

Again we see the Crucial Ballistix Elite pull ahead by a margin of 9.46% in Integer, and 10.28% in Float. That is a consistent margin between the first two tests.

Passmark Performance Test v9.0

Passmark has a variety of test suites, not unlike Aida 64 and SiSoft Sandra. We are only interested in the memory performance, comparing uncached read, cached read, and write tests.

Crucial-Ballistix-Elite-Performance-Test-v9

Passmark shows a bit of a different story between the two kits of RAM. With only 5.11% difference between uncached read, 0.47% in cached read, and 6.33% in write performance, Crucial can only claim a slight victory in the uncached read and write categories, while the cached read is a wash.

CPU-z 1.76.0

CPU-z is a widely known application for retrieving mainly CPU specs, cache, and motherboard information. We are interested in the benchmark for single and muti-threaded performance for this article.

Crucial-Ballistix-Elite-CPU-z-Benchmark

The reference Intel 6700K scores 2064 in single threaded, and 8767 in multi-threaded performance. Comparing these numbers it’s pretty much a wash. The biggest difference here is -0.75% in multi-threaded performance to the Corsair Vengeance kit. This was consistent when tested multiple times. I’m guessing the lower CAS timings play a bigger part in the multi-threaded performance for this test, as memory speed seems to make no difference between these two memory kits.

Testing and results continued on the next page.


SKIP TO PAGE:

<< PREVIOUS            NEXT >>

2 comments

  1. Olle P

    Overall a nice review!

    I use the “Tactical” version, with a more slender heatsink, myself (in a 4x4GB setup).
    When I got the new hardware this summer (around the same time this article was published) I ran similar tests, comparing different settings XMP on/off and also with variations in CPU speed settings.
    The main questions for me at the time were:
    1. Given that Intel officially states support for memory “up to” (only) 2,133 MHz; will faster memory work well at all, or will there be issues?
    2. Given that there are no issues, will faster memory increase the overall computer performance?
    My results matched the ones in this review: No issues. Faster memory will provide a performance increase that’s measurable in benchmarks but hardly notable in regular use.

    Back to your review:
    * The “Elite” heatsink is quite wide. To me it seems like there could be an issue with cooling when all four memory slots are occupied. Is that observation correct?
    * Based on the introductory statement, “In this article for Benchmark Reviews, I’m going to see if higher bandwidth memory makes a difference.”, I would have preferred the Corsair memory with XMP off (or any other memory at 2,133MHz) to be the opposition.
    * It’s refreshing to read a review where the emphasis isn’t on the hardware’s overclocking capabilities!

    Conclusions and recommendations:
    Is Ballistix Elite a good set of RAM? Yes, at least unless you want/need a set of four to run at maximum overclock.
    Generally, when buying RAM the considerations should be (in order of importance):
    1. A type that is compatible with the motherboard and CPU.
    2. Sufficient amount of memory. As of today 8GB is generally considered the minimum requirement (when used with a 64-bit OS) and up to 16GB can be beneficial. For video editing and other demanding tasks even more can be useful.
    3. Do buy RAM in matched pairs to get the most out of the “dual channel” capability.
    4a. For best price/performance buy the fastest RAM you can get at the same (or better) price as slower RAM. Don’t spend extra money to get faster RAM unless you want higher benchmark scores and/or better overclocking capabilities.
    4b. Looks (of the heatsink) is important to some users. Spending extra money to get the right look is fine as long as it doesn’t impede points 1 through 3 above.

    1. Jason Maxfield

      Olle P, There isn’t any cooling issue that I’m aware of when using 4 sticks of this RAM. My friend has 4 installed and we haven’t noticed any heat issues with it so far.

      Memory speed can increase performance a slight amount. But it largely depends on what application is being used. Video or image editing is usually more effected by the amount of RAM installed and not so much the speed. I did notice a slight improvement in my testing with gaming. Although, it’s mostly insignificant. You would not be able to tell simply by observing the game while playing.

      They are good sticks of RAM. They were slightly better than the Corsair RAM I put them up against in the article. Basically you have it right. What your PC is going to be used for will determine how much RAM you will need. Gaming, I’d recommend 16GB. Some games are becoming huge resource hogs and the extra RAM now will help in the future. If you do a lot of video editing, then more is even better.

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>

CAPTCHA

*