G.Skill Trident Z RGB 32GB DDR4-3600 Review

I’m not much for long product names, but Trident Z RGB doesn’t seem too excessive. Add the DDR4-3600 to it, and it’s getting slightly lengthy. Oh, and there’s the CAS 16 rating. Fortunately, G.Skill makes its model numbers a little easier to read than many of its competitors, as the “4” stands for DDR4, 3600 is the data rate, C16 is the latency, the Q means quad-DIMM, 32 is the gigabytes of capacity, and GTZ is the style. Unfortunately, R stands for RGB on its black-and-gray heat spreaders, where G.Skill uses the same letter on other modules to denote the color red. And with that, model F4-3600C16Q-32GTZR is born.

So much for the decoder ring. On to the details.

Specifications

Just any CAS 16 timings wouldn’t do for G.Skill, as the firm understands 16-16-16 is quicker than the 16-18-18 latency found on certain competing modules. It’s even a hair quicker than the classic DDR3-1600 CAS 7 standard, since (quick math) 3200/16*7=1575. Remember that latency is quoted in cycles, and that cycle time is the inverse of frequency: Faster cycles occur in less time.

Enthusiast are also aware that four-DIMM sets can be used in both quad-channel and dual-channel configuration. Oddly, the box includes two dual-DIMM packages, each with its own case badge. One side of the module is black, the other gray, and the black side faces outward on most motherboards.

The white part is of course a plastic light diffuser for the onboard RGB LEDs, which G.Skill said can be controlled by Asus’ Aura application. A quick test of that claim proved that the lights of each module can be controlled individually, or they can be grouped, and that patterns such as Rainbow will assign various colors to the four LEDs of an individual module.

G.Skill's own version of the software, Trident Z RGB Control, is currently in beta. Tested on a non-Asus motherboard, the Static, Breathing, and Strobing options dimmed the LEDs entirely (see solutions in reader comments below). Color cycle, Rainbow, Comet, Flash and Dash, Wave, Glowing Yoyo, Starry-Night, and Music all functioned as advertised.

In addition to the Aura program’s general lighting patterns, the LEDs can also be programed to change color in response to heat. G.Skill’s promotional photo shows just one of several lighting options.

Most motherboards will boot this DDR4-3600 CAS 16 at DDR4-2133 CAS 15 prior to initiating the motherboard’s XMP setting. After that, the modules will boot at DDR4-3600, but only on boards and processors that can reach DDR4-3600. We recommend reading reviews of your board to determine its limits before choosing high-speed memory.

Comparison Products

Test System

Our latest memory test system exploits the stability of Asus’s ROG Maximus IX Hero to leverage the strong overclocking capability of Intel’s Core i7-7700K and the graphics bottleneck-smashing MSI GTX 1080 Armor O/C. With system bottlenecks mostly eliminated, the most direct comparison might appear to be Team Group’s XTREEM DDR4-3600. That’s despite the capacity difference, since our benchmarks don’t need more than 16GB total memory to reach peak performance.

But this still wouldn’t be a fair fight, because the G.Skill Trident Z RGB kit we received has more modules. A test of Team Group’s own Dark ROG DDR4-3000 dual-rank DIMMs revealed the magnitude of the Core i7-7700K’s preference for at least four ranks of RAM, and the T-Force XTREEM contains two single-rank DIMMs. The T-Force Dark ROG got to four ranks by having two ranks on each of its two modules, but it’s only DDR4-3000. The only four-rank kit from our previous reviews that’s fast enough to put a total performance perspective on the Trident Z RGB DDR4-3600 is G.Skill’s own Trident Z DDR4-3866.

MORE: Best Memory

MORE: DDR DRAM FAQs And Troubleshooting Guide

MORE: All Memory Content

Create a new thread in the Reviews comments forum about this subject
This thread is closed for comments
25 comments
Comment from the forums
    Your comment
  • drwho1
    Cons:
    Expensive
    It has Led Color.... should also be a Cons.
    1
  • g-unit1111
    That price. Ouch. But I would be curious to see how the extra speed is handled in a Ryzen rig vs an Intel i7.
    0
  • MASOUTH
    "...where G.Skill uses the same letter on other modules to donate the color red."

    denote?
    3
  • Jack_Burton
    It was definitely worth the extra $40 or so dollars I got my sticks for to me. These serve as the only real LED lighting going on in my case other than some other brand badges that light up white on the cooler and vid card. Looks pretty damn sick set to AMD red. Minimal.
    -1
  • Adm1ra1P
    Anonymous said:
    "...where G.Skill uses the same letter on other modules to donate the color red."

    denote?

    Damn you auto-correct?
    0
  • BulkZerker
    Anyone winging about the LED lighting needs to grow up.
    You can buy sticks without the lighting for less, much less, if not having LEDs is so important to you.
    0
  • chumly
    @hardware_tom
    There are two competing motherboard chipsets right now, and consumers that care as much about the one as the other, you guys should be testing both at this point otherwise you are ignoring the market. You guys planning on testing AMD and Intel separately? Surely the science behind performance would change drastically given a complete change in platform. The results will be worthless to anyone that decides to buy into the other market.
    4
  • GoldMan27
    Anonymous said:
    That price. Ouch. But I would be curious to see how the extra speed is handled in a Ryzen rig vs an Intel i7.

    Same. I'm running a Ryzen 7 1700 with Corsair Vengeance LPX 16GB DDR4-2800, but I could only imagine how much faster my build would be with that.

    Then again, I'm not sure if my board could handle it. ASUS' website says my board (ROG Crosshair VI HERO) can handle up to 64GB DDR4-3200 on OC, so it might not be able to handle the load. Though who am I kidding? I don't have enough money to blow on this just for fancy lights. :P
    2
  • g-unit1111
    Anonymous said:
    Anonymous said:
    That price. Ouch. But I would be curious to see how the extra speed is handled in a Ryzen rig vs an Intel i7.

    Same. I'm running a Ryzen 7 1700 with Corsair Vengeance LPX 16GB DDR4-2800, but I could only imagine how much faster my build would be with that.

    Then again, I'm not sure if my board could handle it. ASUS' website says my board (ROG Crosshair VI HERO) can handle up to 64GB DDR4-3200 on OC, so it might not be able to handle the load. Though who am I kidding? I don't have enough money to blow on this just for fancy lights. :P


    Yeah same. :lol:

    But that is interesting though - even the highest end AM4 boards - the Crosshair VI and the MSI Xpower Titanium - only support max DDR4-3200. I wonder if it's just limitations of the BIOS that could be fixed with a future update or is it limited by the CPU's memory controller or something else?
    0
  • ravewulf
    A few small corrections to the RGB section:

    "and that patterns such as Rainbow will assign various colors to the four LEDs of an individual module."
    it's five LEDs per module, not four.

    I'm running a dual channel kit on a Gigabyte GA-AX370-Gaming K5 using static mode and it works (the other ones you cited as not working may also work, but I haven't tested them). The trick is that there's a bug in the software where the LED brightness is turned all the way down by default even though the color wheel shows differently. It's listed under the known issues on the G.Skill forum

    Quote:
    There's a bug where the default color has its brightness turned all the way down, even though it shows up as red. So the small circle inside the triangle is not on the lower left corner. And when you change the color by rotating the knob on the color wheel, the brightness value doesn't change. The best way to resolve this (if you are on first install or after clicking on "Default") is to click the red, green, or blue box to the left of the color wheel. Then click the button below the five color group selectors to apply the same color to all groups. After this step, you can move that color wheel to any color you want for any color group without the brightness issue. (Note: The other lighting effects, like comet and flash and dash ignores that brightness value, so that color bug doesn't affect those lighting effects.)


    http://www.gskill.us/forum/showthread.php?s=e7362d5bdfbea3f1fd8fbcb5cb066d37&t=14257
    1
  • Crashman
    Anonymous said:
    @hardware_tom
    There are two competing motherboard chipsets right now, and consumers that care as much about the one as the other, you guys should be testing both at this point otherwise you are ignoring the market. You guys planning on testing AMD and Intel separately? Surely the science behind performance would change drastically given a complete change in platform. The results will be worthless to anyone that decides to buy into the other market.

    1. No.
    2. Jacob is working on an AMD memory article. It has a slightly different focus unique to that platform's benefits.
    3. Maybe.
    4. Probably.

    Anonymous said:
    A few small corrections to the RGB section: it's five LEDs per module, not four.

    I'm running a dual channel kit on a Gigabyte GA-AX370-Gaming K5 using static mode and it works (the other ones you cited as not working may also work, but I haven't tested them). The trick is that there's a bug in the software where the LED brightness is turned all the way down by default even though the color wheel shows differently. It's listed under the known issues on the G.Skill forum

    Quote:
    There's a bug where the default color has its brightness turned all the way down, even though it shows up as red. So the small circle inside the triangle is not on the lower left corner. And when you change the color by rotating the knob on the color wheel, the brightness value doesn't change. The best way to resolve this (if you are on first install or after clicking on "Default") is to click the red, green, or blue box to the left of the color wheel. Then click the button below the five color group selectors to apply the same color to all groups. After this step, you can move that color wheel to any color you want for any color group without the brightness issue. (Note: The other lighting effects, like comet and flash and dash ignores that brightness value, so that color bug doesn't affect those lighting effects.)


    http://www.gskill.us/forum/showthread.php?s=e7362d5bdfbea3f1fd8fbcb5cb066d37&t=14257


    Thanks. Considering the PITA of the bug, I'll just leave it as "broken" and put an (see comments below for solutions) in the article.
    0
  • CeltPC
    I am surprised that there is no mention of the design / software flaws of the G.Skill RGB memory modules in this review. The use of controlling software such as G.Skill's or Asus Aura has been corrupting the SPD's for many users. This of course causes major stability issues, and the problem tends to be progressive, getting to the point of being inoperable.

    These problems have been well documented on many forums, including the G.Skill forum. At this point G.Skill and Asus have stated they are working on some kind of solution, but until / if that happens, any owner of these modules are putting their sticks at risk if color controlling software is used.

    I would advise any owner to run an SPD check for corruption using something like Thaiphoon Burner.
    1
  • Bsquared
    I, too, am surprised at the cost premium. Didn't Gskill used to stand for quality memory at affordable prices? Are there any cost-effective RGB DIMMs out there?
    0
  • IceMyth
    Why G.Skill is not benchmarked against Corsair?
    0
  • Crashman
    Anonymous said:
    Why G.Skill is not benchmarked against Corsair?
    Because it was benchmarked against kits with the same number of ranks. The most recently-tested Corsair modules were single-rank two-DIMM kits. It's all explained in the article.
    0
  • Mike Zimmerman
    Check SPD CRC of your memory modules! I bet you have corrupted it.
    0
  • csdani84
    The quick math calculation on latency seems to be wrong.
    It says "3200/16*7=1575". It is rather 3600/16*7=1575. And it means it is a bit slower than DDR3-1600 CAS 7, since its speed equals a theoretical DDR3-1575 CAS 7. Or see the other way around: 1600/7*16=3657 and some, so the DDR3 would equal a theoretical DDR4-3657 CAS 16 regarding CAS latencies.
    0
  • nbrikha
    So, just no point on testing RAM on an AMD rig?
    0
  • Nintendork
    They should be testing on Ryzen only and ditch intel for good. Ryzen is an arch that actually benefits from fast ram, perfect excuse for high speed memory as a reviewer.
    0
  • Crashman
    Anonymous said:
    They should be testing on Ryzen only and ditch intel for good. Ryzen is an arch that actually benefits from fast ram, perfect excuse for high speed memory as a reviewer.
    So you have an AMD platform that will test DDR4-4200 at full data rate? That's excellent, please send it :)

    The focus of these articles is to see how far the memory can be pushed. If you want to see how much performance can be had without pushing the limits of the memory, our AMD guy has an article in the works for you. And then you'll see two types of memory articles, CPU focused and memory focused. And the memory focused one will still use whatever platform supports the highest data rate.

    Now I get that people have grown impatient waiting for the AMD article, but the guy working on it is a true enthusiast with a full-time job elsewhere. I'm not going to push him to rush out a half-hearted article, and I'm not going to stress him to the point where he gives up and sells the late article to another publisher. I'm going to be nice to the guy and allow him the time he needs to complete the article to his own satisfaction, because he deserves that consideration.
    1