Home Editorial Kingston HyperX Beast review: Is this RAM the best?

When Kingston originally announced the release of its HyperX Beast RAM, I was skeptical. Sure, the memory looked sleek enough,* thanks to its black PCB and sexy heat spreaders… but what could HyperX DDR3 RAM offer me that my current GSkill Ripjaws X sticks couldn’t?

Instead of leaving it to chance, I decided to take Kingston up on their review offer and pit these competing brands of RAM against one another, memorio a memorio. As a relatively heavy “power user,” I’d be pushing both sets through gaming, video production, and (of course) benchmarking. Delicious.

Appearance: black is the new black

While Kingston’s strong focus on the color of its new RAM wasn’t something that bowled me over right away,** I have to admit HyperX Beast memory is extraordinarily seductive, as far as PC gaming hardware goes. Each stick’s matte black finish and muscular curves immediately reminded me of the classic 1989 Batman, which is certainly no complaint. There’s something about HyperX RAM that makes you want to touch it, which is a shame, given that it’s going to spend 99.5% of its time locked away in your PC’s case.

The heat spreaders on each Beast stick are quite tall, almost a full centimeter over my GSkill Ripjaws X sticks, which are already sizeable, to say the last. While this could be a problem in a mini tower or a particularly crowded case, I found it made HyperX Beast memory easy to install quickly and with confidence. Mind you, installing memory isn’t exactly hard, but anything that minimizes the nerve-wracking snap of PC slot installation is always a plus in my book.

Whether the heat spreaders are as functional as they are Tim Burton-esque is hard to say, as I’ve never truly had temperature problems in my custom case. But these Kingston modules certainly didn’t change the status quo for the worse, so at the very least they’re competent at keeping temps within a reasonable range, even under heavy load.

kingston hyperx ram

Speed: ramming towards the finish line

Testing 8GB of Kingston HyperX Beast RAM against 8GB of GSkill Ripjaws X RAM (DDR3 1600 for both), I decided my first tasks would be speed-related. I completed a series of trials with on configuration, took averages, then swapped modules and repeated the trial. While both memory manufacturers encourage overclocking, I thought it would be more fair to compare what you get from each option, stock. It’s certainly possible the results could be swayed with some OC loving, but it’s also possible I’d inadvertently set my house on fire attempting it.

Daily heavy usage tests: There was no discernible difference between HyperX Beast and Ripjaws X memory. Both configurations handled pretty much everything I threw at them, with massive amounts of media playing and programs running simultaneously. In all instances, my CPU (i5 3570K) bottlenecked far before my memory did.

Video rendering: Since media production is always  a resource hog, I next pitted Kingston’s RAM against its competitor in a bout of video rendering. With no other background applications running (other than those necessary for Windows to function), I exported a somewhat lengthy Duels of the Planeswalkers video with both configurations.

While the GSkill Ripjaws X RAM finished the task in an average of 47 minutes and 47 seconds, HyperX Beast took an average of 48 minutes and 16 seconds. That comes to about a 1% loss on Kingston’s side, though I can’t say that difference is particularly significant.

Time to reboot: With my ultra-fast solid state hard drive, I was unable to measure any real difference between either RAM setup from shutdown to Start Screen. A full restart took between eleven and twelve seconds with Kingston and GSkill. Again, a fairly inconclusive result.

Kingston HyperX Beast black RAM

Benchmarks: HyperX vs. heavy lifting

Unable to establish a clear winner with my personally-derived assessments, I turned to Passmark Performance benchmarking tools to try to establish a clear winner. It became immediately obvious that, while the differences were slight, my existing GSkill Ripjaws X memory was slightly faster than HyperX Beast RAM of the same type.

Here are the average results (GSkill Ripjaws X and Kingston HyperX Beast, respectively):

  • Database operations: 104,600/second vs 97,700/second
  • Cached read speed: 27,259 mb/s vs 27,201 mb/s
  • Uncached read speed: 15,574 mb/s vs 14,434 mb/s
  • Memory write speed: 10,316 mb/s vs 8,699 mb/s
  • Memory threaded: 23,513 mb/s vs 19,218 mb/s
  • Latency (lower is better): 20.6 nano sec vs 24.5 nano sec
  • Memory Mark composite: 2614 vs 2525

While my own tests indicated a negligible advantage of Ripjaws X memory over HyperX Beast memory, the benchmarks granted GSkill’s RAM a slightly more conclusive edge with an overall Memory Mark composite score 3.5% higher than that of Kingston’s DDR3 sticks.

Is that enough of a difference to completely discredit HyperX Beast as a viable option for your PC gaming dollars?

It depends. These hardware options were clearly still in close competition, and your results may vary based on your unique PC build or usage. If you’re looking to overclock your RAM, you may find Kingston memory outperforms GSkill in that arena, since it wasn’t part of my tests. It’s also entirely possible that a 3.5% difference simply isn’t statistically significant, given the relatively narrow scope of this review.

Ultimately, I don’t think you can go wrong with either RAM option. All things being equal, both GSkill and Kingston modules will offer fast, low temperature performance under heavy load, and both memory modules look seductive inside a gaming PC rig. Go with your gut on the aesthetics or, failing that, simply take whichever costs less on NewEgg when you’re shopping!

 

* Because we all know that counts.
** I believe memory should be judge on the content of its timings, not the color of its heat sinks.

 

2 replies to this post
  1. Hi!, Everything on mine is doubled here,

    Database operations: 104,600/second vs 97,700/second
    Cached read speed: 27,259 mb/s vs 27,201 mb/s
    Uncached read speed: 15,574 mb/s vs 14,434 mb/s
    Memory write speed: 10,316 mb/s vs 8,699 mb/s
    Memory threaded: 23,513 mb/s vs 19,218 mb/s
    Latency (lower is better): 20.6 nano sec vs 24.5 nano sec
    Memory Mark composite: 2614 vs 2525

    But my latency is also doubled? Its currently at 46.3

    Can you help me fix that?

Leave a Reply