Table of Contents
Team Group recently released its answer to the ever-growing problem of XMP memory compatibility on AMD X570 series motherboards. Deriving it’s name in part from the Greek alphabet, the new T-Force Dark Zα, or Z-Alpha is poised to make a big splash in the AMD memory market. Team Group’s marketing assures that this memory has been tailor-made for the current generation AMD platform. We are going to put that claim to the test, as well as see what this memory has to offer in terms of overclocking headroom. Come join us as we explore this memory and see what it has to offer.
Specifications and Features
There are many factors to consider when purchasing system memory. Most people end up focusing on the physical characteristics and how the memory will look in their system. While those are important elements to choosing memory, the biggest difference in price often comes down to the XMP rating and effective speed. There can be as much as a 75% price difference for memory within the same lineup.The biggest difference in pricing comes down to the actual ICs used in the modules. Higher speeds with tight timings will always have higher pricing as only certain ICs can achieve this.
Purchasing memory for an AMD X570 series build can be complicated. Many of the memory kits sold on the market today are intended for Intel-based chipsets. As modules designed for Intel chipsets run slightly different timing profiles, this can have an adverse effect on an AMD platform, and in some cases, the XMP profile simply won’t boot. Thankfully, Team Group makes choosing memory easy, because every Dark Zα memory module has been specifically built for X570 motherboards.
Team Group offers the Dark Zα in six different configurations. There are three different XMP speeds to choose from, all of which come in either 2 x 8GB or 2 x 16GB densities. The lineup includes 3200 CL16, 3600 CL18, and at the high-end 4000 CL18. All of the memory available in this lineup has been strenuously tested to be compatible with AMD Ryzen 3 series motherboards and CPUs.
Today, we will be evaluating the 3600 CL18 memory kit. In the table below, we examine the particular details of our test memory.
|T-Force Dark Zα DDR4 Gaming Memory for AMD|
|Capacity||16GB (8GB x 2)|
|Speed Spec||PC4 28800|
|Rated Frequency||3600 MHz|
For those interested, here is a closer look at the particular details of this kit of memory. Below is a screenshot of Thaiphoon Burner, which is a wonderful free tool that allows one to read the Serial Presence Detect (SPD) firmware of the DRAM. The SPD information is critical in determining how the stick will perform and how the computer will recognize it out of the box.
As the Thaiphoon Burner screenshot shows, this specific kit of memory is composed of SK Hynix C-Die ICs.
It’s not easy to spot, but a sharp eye will see that this memory utilizes the modern A2 style PCB layout. This refers to the orientation of the 8 ICs on the actual memory PCB itself. While it may seem like a useless tidbit of information, this is relevant, because it can play a very substantial role in the relative overclocking ability on some motherboards.
Packaging and Product Tour
Team Group ships out the new Dark Zα memory in plastic clamshell-style packaging which we often see on lower-end products. While it may be simplistic, it does protect the memory during shipping and provides an opportunity for a sneak peek. This is also a single-use type of packaging. As a result, most users will end up throwing this packaging away once the product is installed in the system, unlike the packaging of a high-end graphics card.
Inside you will find the customary T-Force case badge sticker and a small quick-start manual.
Once out of the packaging, we get our first impression of the build quality. Right away we can see Team Group went with aggressive styling and a shiny black finish. The fins on the top will catch system airflow and help cool the memory. Overall, the heat sink does feel a little thin, and the weight in your hand reflects this as well.
As we detailed above, this memory is running Hynix ICs. We know from prior experience with Hynix DDR4 that the ICs themselves don’t tend to put out much heat at the XMP-rated voltage, The effectiveness of the cooling method remains to be seen.
The heat sink design is a little on the tall side, but it should clear most air coolers. It’s always a good idea to verify that this memory will work with your CPU cooling method before purchasing. The full dimensions are as follows (L x W x H): 141 mm x 43.5 mm x 8.3 mm. The total weight for one complete memory module is 40 grams.
Take a look below for more pictures.
The memory is right at home when paired with the ASRock X570 Taichi motherboard. The armor and heat sinks on the Taichi are more of a matte finish and doesn’t perfectly match the Dark Zα, but we think it still makes a striking assembly.
Testing and Overclocking
AMD introduced serious changes to the memory capabilities and overclocking of the new X570 series chipset. As we talked about earlier, many of the existing XMP profiles on the market today are simply not compatible with Ryzen 3000 series and will not even boot when applied. It’s because of this that our first goal is to check and see if this memory even works with the XMP profile out of the box. We will attempt to apply the XMP profile and run benchmarking software to assess the relative stability.
If and when the XMP profile has been established to be stable, we will evaluate the memory from an overclocking perspective. We want to see what this memory can do, but without hurting it. Therefore, we will stick to what could be classified as 24/7 stable daily memory voltages. According to the XMP 2.0 certifications, the absolute maximum allowable voltage is 1.50 V VDDR. Thus, all overclocking endeavors will be conducted with no more than 1.50 V.
Below is the test system and resulting memory speeds that will be used to evaluate the memory and run the benchmarks.
|CPU||AMD RYZEN 7 3700X @ 4.3 GHz|
|Cooler||NZXT Kraken X62 280mm – All-In-One Cooler|
|Motherboard||ASRock X570 TAICHI AM4|
|Graphics Card||PowerColor RED DEVIL Radeon RX 580|
|Solid State Drive||Team Group L5 LITE 3D SSD|
|Power Supply||Enermax RevoBron 700W 80+ Bronze|
|Operating System||Windows 10 x64|
|Memory Speeds Compared|
|XMP ~ 3600 MHz CL18-22-22 + XMP Sub Timings @ 1.35 V|
|Test Case 1 ~ 4200 MHz CL18-22-22 + XMP Sub Timings @ 1.45 V|
|Test Case 2 ~ 3600 Mhz CL16-19-19 + Tight Sub Timings @ 1.45 V|
|Test Case 3 ~ 4200 MHz CL18-22-22 + Tight Sub Timings @ 1.45 V|
As is the case with all overclocking adventures, your results may vary, so proceed only if you assume all risk. To view and examine all of the various memory profiles, we use two primary tools which include AIDA64 and AMD Ryzen Master utility. AIDA64 is a powerful system diagnostic and benchmarking tool that can be purchased for a reasonable price.
The modern DDR4 market offers a vast range of XMP memory. Shopping on Newegg, we find everything from DDR4 2400 MHz, all the way up to DDR4 4866 MHz. The improvement in XMP profile speeds is greatly attributed to modern manufacturing processes and memory PCB layouts; however, it might not be attainable on all motherboards. The profile is intended to be a one-click overclock, but it likely only applies to enthusiast-grade motherboards.
Crucial ?? designed the Ballistix Elite to work no matter which platform you are running. Whether it’s the new AMD Ryzen 3000 series or Intel’s current Z390, you can buy with confidence, because this memory has been extensively tested with all platforms. That being said, some budget motherboards aren’t designed for high-frequency DDR4. It’s always a good idea to look at the QVL and check with your motherboard manufacturer to make sure 3600 MHz DDR4 is supported.
As you can see below, we had no problem running our memory using the one-click XMP memory profile. In this case, the FCLK = 1800 MHz and UCLK = 1800 MHz.
We cannot examine the overclocking potential without at least briefly covering the new AMD memory structure. Within the new architecture, AMD decoupled the infinity data fabric clock (FCLK) and the unified memory controller clock (UCLK). In the previous generation, the clocks were inherently linked with memory frequency and dividers. The process of decoupling meant that AMD could push the frequency to substantially higher limits, from about 4000 MHz to over 5000 MHz.
When we go to test the memory and overclocking capabilities we need to take into consideration two new variables: FCLK and UCLK. Even though these variables are somewhat out of our control, they play a huge role in the overall performance and benchmark scoring. That being said, it’s not always fair to compare one frequency against another, because the FCLK and UCLK might be running at different speeds and will, thus, skew the test results. While it’s not within the scope of this memory review to dive deeper into the Ryzen 3 memory structure, just keep in mind that those variables are there.
For the first test case only, the operating frequency was increased with no other settings adjusted. With a small voltage increase, the frequency was able to be increased by an astonishing 600MHz. Overclocking mileage and headroom will vary depending on your motherboard, but effectively we achieved a nice overclock on the frequency alone. We should point out that the primary timings are quite loose at CL18-22-22, so substantial frequency overclocks were expected here. Below is the resulting first overclock test case with FCLK = 1800 MHz and UCLK = 1050 MHz.
For the second overclock test case, we used a voltage increase of 1.35 V to 1.45 V. Contrary to the first test case, this time only the memory timings were decreased in order to provide an alternative example of what types of overclocks are possible. The operating frequency was held at the XMP 3600 MHz, but the primary timings were decreased as far as possible while still maintaining relative stability.
It’s worth noting that Hynix’s memory seemed to have difficulties running low primary and secondary timings. This is one area in which Samsung has a clear advantage because B-Die memory modules are able to run at substantially lower timings all around.
Furthermore, we found that Hynix was even more difficult to overclock that Micron memory. Given the frequency limit of 3600MHz, we found it impossible to run primary timings lower than CL16-19-19, which would be considered sub-par by when compared to Samsung memory modules.
Below is the resulting second overclock test case with FCLK = 1800 MHz and UCLK = 1800 MHz.
For the third and final overclocking test case, the goal was to achieve a well-rounded overclocking profile that includes both tighter timings and higher frequency. For this round, we will stay within the maximum allowable voltage of 1.50 V, according to XMP 2.0 specifications.
Our Team Group Dark Zα DDR4 Gaming Memory proved to be difficult to overclock. We were able to achieve 4200 MHz quite easily, but lowering the timings was difficult on the memory. The primary timings needed to be at XMP specifications for all frequencies of 4000 MHz and above, so we didn’t gain any performance there. However, we were able to tighten up the sub-timings. This has proven to be extremely beneficial for performance in the past. It comes down to just a few timings that make the biggest difference.
Adding voltage simply didn’t help the situation and we ended up sticking to 1.45 V because adding more simply didn’t gain us anything. Below is the resulting first overclock test case with FCLK = 1800 MHz and UCLK = 1050 MHz.
First up, we used AIDA64 Cache and Memory Benchmark. In the graph below, it is clearly visible that each of the four different memory speeds compared had a noticeable change in the benchmark result, with the exception of memory write performance. Looking at the read performance, we found it interesting that overclocking the frequency resulted in a lower score than stock XMP. This can be attributed to the substantially lower UCLK frequency in that case.
How did our memory overclocking impact the overall performance? In total, our overclocking endeavors didn’t have a huge impact on performance. It comes down to how we interpret the results and which tests we choose to look at. For the read performance, increasing the frequency had a clearly negative effect. Conversely, in copy performance, high frequency was the clear winner. We gained 5% or more from frequency overclocking.
Due to the inverse relationship of overclocking performance, we feel the XMP profile is the best option.
Next up is Geekbench 3, and it has proven itself to be an excellent tool for determining the real-world performance of the system being tested. This type of benchmark is purely 2D calculation-based and there is no graphical processing element, so it’s a great analytical tool to evaluate memory performance.
Here is where things get interesting. Focusing on the overall score, we see that this ‘real world’ test benchmark favors the tighter sub-timings. In both cases where we altered the timings, we were able to increase the performance by at least 4%. We have never seen a trend develop in Geekbench 4 where adjusting one thing helped all 4 tests. Although, that’s exactly the case with this memory. Adjusting the sub-timings had an overall beneficial effect on the performance.
Cinebench R15 scores don’t typically increase much with memory overclocking, but we included it because it’s quite possibly the most popular benchmark for computers today.
WinRAR is an old benchmark, but still quite useful in modern computing. It evaluates performance by simulating file management tasks such as compression and extraction. Here we see a noticeable trend of 3600 MHz, with tight timings being the overall winner.
Next, we will examine the performance using a few of the memory benchmark tests offered within the SiSoftware Sandra suite of benchmarks. The flagship product, known as Sandra, is a powerful suite of many different benchmarks used to evaluate computer performance of all major components, including the processor, graphics, memory, and disk.
As with the AIDA64 memory test results, interpreting the SISandra memory results is tricky because different tests represent different findings for overclocking. The high frequency profiles are clearly better for the Cache & latency test. In the memory bandwidth test, we saw the biggest improvement in score for all benchmarks tested. With a total increase of 15%, test case 3 was the clear winner.
In this next section, we will examine the XMP profile performance of recently-reviewed memory. All of the memory compared here have been tested on the same exact X570 test bench and with the same exact operating system. Due to this fact, we are able to make some direct performance comparisons of recently-reviewed memory. For the purposes of continuity, we will examine only the XMP profile and exclude any overclocked comparisons.
Last month, we reviewed G.Skill’s Trident Z Neo and Crucial’s Ballistix Elite memory. Using the data generated in those reviews, we are able to compare it with the Team Group Dark Zα. All of the comparisons below were done with the XMP profile and disregard any overclocking potential.
As we might expect, the Team Group Dark Zα came in last, because the primary timings are substantially higher than the other memory modules being compared. While they were remarkably close in performance, due to the fact that they are all running 3600 MHz, G.SKill is running the tightest primary timings which give them the edge here.
The story remains the same with Geekbench 4. The result which we should focus on here is the overall test. This gives the best picture of the actual performance which one can expect from the XMP profile. For the overall test, we find that all memory modules compared are strikingly similar.
Our overall impression of Team Group’s T-Force Dark Zα is very positive. From the aggressive but not gaudy styling to the speedy XMP profile, this memory packs in high-end performance at budget pricing. With the stamped aluminum heat sink design and a glossy finish, Team Group delivers elegance and cooling power all in one package.
Team Group chose to outfit the Dark Zα with SK Hynix memory ICs. From an overclocking standpoint, this proved to be a less desirable memory IC option compared to Samsung. We also had difficulties achieving stable frequencies higher than 4200 MHz. When it came to lowering the primary timings the Hynix IC also proved to be a disadvantage. However, If you’ve followed with us up to this point then you have probably come to your own conclusion that overclocking the memory for daily purposes on the X570 platform is not practical. The new XMP profiles built for AMD platforms are extremely efficient and easy to use with no overclocking required.
The real story of this memory is price-per-performance. As we have seen in the comparison above, Team Group’s T-Force Dark Zα is right in line with the competitors in terms of system performance, but at a fraction of the cost. Coming in at a mere $88.99, the Dark Zα has nearly identical XMP performance as G.SKill’s Trident Z Neo priced at $159.99, and Crucial’s Ballistic Elite priced at $199.99. While it might not be the top choice for competitive overclockers, we feel that the Dark Zα is a smart choice for gaming and performance-minded computer builds.
David Miller – mllrkllr88