Last week, we reviewed the Trident Z5 RGB 7200 MT/s CL36 memory kit, and today we’re going to take a look at its big brother, the Trident Z5 RGB 8000 MT/s CL40, also from G.SKill. We’ll see in our article that, in the end, these two kits are very similar. This kit is already available in France and can be found for 336 euros on Amazon, which is a reasonable price compared with the competition. Our sample of the day consists of two 24 GB memory sticks and is equipped with the new Hynix M-die chips, which prove to be quite docile when overclocked.
Don’t hesitate to ask us any questions you may have, either in comments or on our social networks. You can also join our community on the discord. We’d also like to take this opportunity to thank G.SKill for providing us with this copy of the 48 GB Trident Z5 RGB 8000 MT/s CL40 kit.
Here’s an update on the kits available to us:
Since this is an article in which we’ll be doing comparative tests, we thought it would be a good idea to review the various kits we have. You’ll be able to find out more about the features of each one. You can also take a look at the section where our latest memory kit tests are available, and find the one that interests you. Don’t forget that as of today, our test platform is evolving, so we’ll have new performance tables.
- 48 GB Trident Z5 RGB 7200 MT/s CL36 (Hynix M-die)
- 48 GB Patriot Xtreme 5 8200 MT/s CL38 (Hynix M-die)
- 48 GB Corsair Dominator Titanium 8000 MT/s CL38 (Hynix M-die)
- 48 GB Teamgroup T-Force Xtreem 8200 MT/s CL38 (Hynix M-die)
- 32 GB Corsair Dominator Titanium 7600 MT/s CL36 (Hynix A-die)
- 32 GB Corsair Dominator Titanium 7200 MT/s CL34 (Hynix A-die)
- 48 GB Corsair Vengeance RGB 7000 MT/s CL40 (Micron)
- 32 GB GSKill Trident Z5 RGB 6400 MT/s CL32 (Hynix M-die)
- 32 GB Corsair Dominator Platinum RGB 7200 MT/s CL34 (Hynix A-die)
- 32 GB GSKill Trident Z5 RGB 6800 MT/s CL34 (Hynix A-die)
- 32 GB Teamgroup Deltra RGB 7200 MT/s CL34 (Hynix A-die)
- 32 GB Kingston Fury Renegade 6400 MT/s CL32 (Hynix M-die)
- 32 GB Corsair Vengeance RGB 6400 MHz CL38 (Hynix M-die)
- 32 GB Teamgroup Delta RGB 6600 MT/s CL34 (Hynix M-die)
- 32 GB Corsair Dominator Platinum RGB 6600 MHz CL32 (Hynix M-die)
- 16 GB Kingston Fury Beast RGB 6000 MHz CL40 (Hynix M-die)
- 32 GB Corsair Vengeance 5600 MHz CL36 (Samsung)
- 32 GB Teamgroup Delta RGB 6400 MHz CL40 (Hynix M-die)
- 32 GB Dominator Platinum RGB 5600 MHz CL38 (Samsung)
- 32 GB Kingston Fury Beast 4800 – 5200 MHz CL40 (Micron)
- 32 GB GSKill Trident Z5 5600 MHz CL36 (Samsung)
[nextpage title=”Packaging description:”]
The G.SKill kit is delivered in a cardboard box with the usual color code in keeping with the spirit of the brand. You can see the label of our two black bars on the back of the box thanks to the two cut-outs. You can also see that these are enclosed and protected in a transparent plastic cover. On the rear of the box is a label informing future purchasers of the characteristics of their memory kit. Two cut-outs on the labels provide confirmation of this information.
Once the box is opened, we find our two modules perfectly protected in a transparent plastic shell. This “Trident Z5” series arrived with the arrival of DDR5 memory kits.
The Trident Z5 kit is available in either gray or black. What’s more, you’ll have the choice of opting for a version with or without RGB backlighting. Our sample of the day is black with RGB backlighting.
Features and cooling :
Our kit is clocked at 8000 MT/s, has timings of 40-48-48-128 and requires a voltage of 1.35 volts to operate. It consists of two 24 GB modules for a total of 48 GB. This is a mix of matte black and black.
Heatspreaders are made of aluminum. They are responsible for dissipating the heat generated by the eight-layer PCB and the memory chips. If you’ve already had your hands on a G.SKill Trident kit, you’ll quickly realize that G.KSill has kept the same spirit in the heat spreader design.
Both sides feature the series name. A label includes some of our kit’s specifications. In terms of price, this kit is currently available on Amazon for 336 euros. It’s possible to find it even cheaper elsewhere, but you’ll have to look to Germany. On the top, you’ll find the RGB backlight with the G.Skill lettering.
Finally, the kit is XMP 3.0(Extreme Memory Profile) compatible. So don’t forget to activate the profile directly via our motherboards’ bios.
[nextpage title=”How to identify your memory kit:”]
Basic features :
You’re going to tell us that this is pretty easy, since all you have to do is look at the references advertised by the manufacturer. Yes, but that’s not enough, especially if you’re planning to overclock it! Some information is given by the manufacturer, but you’ll have to dig around – or even remove the heat sinks – to find out the rest, or search the web for confirmation.
One side of the heatsink reads “Trident Z5 RGB”, while the other has a label with the technical specifications of our example, which looks more like a code name: “F5-8000J4048F24GX2-TZ5RK”. This is what we’re going to look at in detail. For example, in the case of today’s model, here’s the information we know.
- Brand: G.SKill
- Model: Trident Z5 RGB
- Capacity: 2 x 24 GB
- Frequency: 8000 MT/s
- Timings: 40-48-48-128
- Voltage : 1.35 volts
- Review: –
In order to get an idea of the overclocking potential of your kit, this information is insufficient and you’ll need to go further. It’s also very important to know the revision of the PCB and the model of the chips used in your memory kit. Depending on this information, the kit’s overclocking performance will be very different.
To find this information, there is currently only one solution: disassemble the heat sinks. The Taiphoon Burner software does not currently support DDR5 kits. Nor is there any indication that this will be the case in the future.
CPU-Z always present :
Fortunately, our trusty CPU-Z software is still at work, enabling us to verify this information, particularly with regard to the type of chips used in our kit. At present, there are three types of chip: Samsung, Hynix and Micron.
Our kit is perfectly recognized under Windows 10 and the various timings perfectly read. The memory chips used in our kit are Hynix. These are the chips with the best overclocking potential. As this is a 2 x 24 GB kit and not a 2 x 16 GB, we already know with virtual certainty that these are second-generation Hynix M-die chips.
Our memory kit PCB :
Our kit benefits from 3 GB chips, of which there are 8. We’re using Hynix M-die here, not A-die, but this is a new generation which further improves frequency rise. Our kit is an “SR” model for single rank, meaning that all the chips are on a single side. On the other hand, there are “DR” kits for dual rank , with chips on both sides of the PCB.
[nextpage title=”The importance of the XMP profile:”]
BIOS settings :
We can’t stress this enough, but your memory kit ‘s parameters are not automatically set once you’ve installed it. If this is the case for your processor, it’s not so for your memory kit, and you’ll have to go through the BIOS! This is what we had to do with our Maximus Z790 APEX Encore to get the BIOS to load the XMP profile.
You can then configure it manually, choosing the right frequency, timings and voltage for your kit. For ease of use, we advise you to activate XMP (Intel) or EXPO (AMD) mode in ” AI Overclock Tuner ” to set it up automatically.
Without this, your performance will be impacted in all areas, and you’ll end up with a kit that doesn’t exceed the CL40 frequency of 4800 MHz. Performance will then be extremely limited. So let’s activate our XMP profile directly in the BIOS. The XMP1 proposed is the 8000 MT/s profile. The other two XMP profiles, XMP2 and XMP Tweaked, are very slightly different in terms of subtimings and sometimes perform slightly better.
ROG BIOS, whether on APEX or Crosshair models, benefit from memory profiles. They are a good basis for overclocking your memory kit. We’ll start by using them in our overclocking tests. Our kit is now set up in the BIOS and will work with the manufacturer’s specifications.
As you can see from the screenshots above, once the XMP profile has been activated, the settings are automatically detected. The frequency is set to 8000 MT/s, timings 40-48-48-128 are activated and the voltage is increased to 1.35 volts.
[nextpage title=”Intel configuration test:”]
Of course, this kit will only be tested on an Intel platform, as our comparisons have been based on this configuration from the outset. We’ve naturally opted for the ROG Maximus Z790 APEX motherboard, but we’re also testing the latest refresh model in parallel , the ROG Maximus Z790 APEX Encore version.
- Motherboard: ROG Maximus Z790 APEX ENCORE (BIOS 0080)
- Processor: Intel i9-13900KS
- Memory: 48 GB G.Skill Trident Z5 RGB 8000 MT/s CL40
- Graphics card: MSI GT 710
- SSD: Western Digital SN850 NVMe SSD M.2
- Power supply: be quiet! Dark Power PRO 12 1500w
- Operating system: Windows 10 64-bit (22H2)
We’ve also decided to stick with Windows 10 (22h2), as it’s proving to be a better performer, especially when it comes to overclocking. We’ve noticed a significant drop in processor and memory performance since the latest Windows 11 updates, which sometimes distorts the results. Our benchmarks are run a minimum of three times, and the result is the average of these scores.
Aida 64:
This is a program that can provide you with a wealth of information about your configuration, but above all it has a memory benchmark. This calculates the read, write and copy speeds, as well as the latency, of your memory kit. It is often used to compare memory kits. It’s an excellent tool if you want to compare the performance of your memory kit with those we’ve already tested.
The version used for testing is 6.92.6603 beta, and we’ve used the performance of previously tested kits.
The performance of this first test is exactly what we expected, with a read throughput of 122703 MB/s for our 8000 MT/s XMP profile. The position in this first graph is logical for this frequency. The primary timings are more relaxed, which means that performance lags slightly behind an 8000 MT/s CL38 kit. So, on an Intel platform, you’ll need to prioritize frequency and tight timings.
The same applies to write speeds, which are good, reaching 120750 MB/s. The logic of the graph remains identical to the previous one. The Corsair Dominator Platinum 4800 MT/s CL19 in DDR4 achieved 67251 MB/s. Once again, we note a logical progression between the different kits according to their frequency and timing.
The findings are identical in copy, with a position just behind the 8000 MT/s kit, but in CL38. The Corsair Dominator Platinum 4800 MT/s CL19 in DDR4 achieved a copy speed of 57330 MB/s. As you can see, it’s in reading and copying that DDR5 takes a huge lead over DDR4. Note also the difference between a DDR5 kit at 4800 MT/s (72415 MB/s) and our kit at 8000 MT/s (117276 MT/s). If we hadn’t activated our XMP profile in the BIOS, our kit, although sold as an 8000 MT/s model, would really have been clocked at 4800 MT/s. A reminder of the importance of enabling the XMP profile in the BIOS.
Finally, latency. This is the value on which all eyes are focused, as DDR5 lags a little behind. Our best DDR4 kit in this respect was the G.SKill Trident Z Royal Elite 4000 MT/S CL14, which achieved 41.4 ns. You’ll notice that the higher the DDR5 frequency, the narrower the gap. Note that a 3200 MT/S CL16 DDR4 kit has a latency of 50.3 ns. Here, our G.Skill Trident Z5 RGB 8000 MT/s CL40 memory kit has a latency of 56 ns.
Geekbench 3.4.4:
This benchmark is available in several revisions, version 5 having just been released a few weeks ago. It provides two scores: one single and one multi. Using this benchmark, we’re going to focus on the performance obtained on the ” Memory Multi-Core ” score, which differentiates memory performance. The version used for testing is 3.4.4.
For the Geekbench 3 benchmark, despite the single-core frequency of our i9-13900KS being 6 GHz, there’s a difference in score between the different memory kits. Our sample of the day occupies sixth place, even though the difference is almost non-existent.
In multicore, we’re always at the top of the graph, with kits with higher frequencies in front, and vice versa, for the last places.
As soon as we isolate the memory score, real differences are visible between the kits. Our kit of the day scores 13827 points! We’ll be keeping a close eye on this score when we start overclocking. It’s a good indicator of the benefits of overclocking.
In DDR4, our best kit scored just 7748 points. Once again, the enormous advantage of DDR5 combined with the Raptor Lake processor.
Geekbench 5.5.1:
We’ve decided to add this extra benchmark, but why? The CPU benchmark uses new tests that more closely simulate the tasks facing processors in the face of recent applications. Geekbench 5 also increases the memory used in the benchmark to better reflect the impact of this parameter on CPU results. For the time being, we’ll be keeping both, as version 3 is still widely used. We have also migrated to the latest version, 5.5.1.
Whereas in Geekbench 3, we had a difference in single core, here in version 5, the gaps are smaller but, as with the previous version, we find a similar ranking.
Finally, the score for our G.Skill Trident Z5 RGB 8000 MT/s CL40 kit is 28434 points. Still to give you a reference score in DDR4, our best score in Geekbench 5 was 11,179 points.
Geekbench 6:
We recently received a key for Geekbench version 6. We therefore took advantage of the change of configuration to add this new benchmark to our test suite. This is the latest benchmark and is intended to be more representative of current configurations. Primate Labs also assures us that the scores displayed by Geekbench 6 are more accurate in relation to the various existing devices and platforms. In fact, the test is much longer than the other two versions we use. Our tests will be carried out on the latest version, i.e. 6.2.2.
The scores in the single test are fairly close, which explains the small but insignificant change in order.
Let’s conclude this series of benchmarks with the multi part of Geekbench 6. Logic is once again respected, which means that on the Intel platform, you need to prioritize frequency and tight timings for maximum performance.
[nextpage title=”Overclocking on our APEX Z790 Encore:”]
Choosing your DDR5 kit according to memory chips :
As was already the case with DDR4, if you want to clock your memory, certain brands will benefit from better potential. At present, the kits announced by the various brands benefit from either Micron, Hynix or Samsung chips. The latter two brands will be the most interesting. As you can see, our G.SKill Trident Z5 RGB kit, clocked at 8000 MT/s CL40, is equipped with Hynix chips. Unlike our last kits, these are not Hynix A-die chips, but the new Hynix M-die chips. These have the particularity of rising more easily in frequency, but being more complicated to tighten in terms of timing.
To take advantage of high-frequency kits, it’s best to be equipped with a motherboard dedicated to this type of kit, and in this game, the ROG APEX Z790 is certainly the best. It features two memory slots to facilitate memory overclocking. You’ll also need a processor with an excellent IMC(Integred Memory Controller). Beware: a processor may have excellent overclocking potential but, on the other hand, a disappointing IMC. These two factors are independent of each other.
Key DDR5 voltages :
For memory overclocking, there are 5 key voltages. These are the voltages that will have an impact on overclocking and therefore on frequency rise or timing tightness. These are the CPU System Agent Voltage (SA), the DRAM VVD Voltage, the DRAM VVDQ Voltage, the Memory Controller Voltage and the IVR Transmitter VVDQ Voltage. Here’s where you’ll find these voltages in the BIOS of our ROG Maximus Z690/Z790 APEX.
Please note that, depending on the type of memory chip (Micron, Hynix and Samsung), the balance between these different voltages is different. If your PMIC module is not locked, by activating ” High DRAM Voltage Mode “, you’ll have access to voltages up to 2.00 volts. Now it’s time to overclock this G.SKill DDR5.
Step 1: use available BIOS profiles
It’s a strange thing to say, but a few months ago, we were happy to have the BIOS profiles available as a good basis for our work. However, since the arrival of the Hynix A-die and now the new Hynix M-die chips, the frequency of kits has exploded, and in the end, the profiles in the BIOS are lower than the current frequency of our kit. There’s a 7600 MT/s CL36 profile and an 8800 MT/s profile for Hynix A-die chips, so we’ll have to build our own profiles.
Step 2: 8800 MT/s CL36 target
This is our second objective: to reach the 8800 MT/s CL36 frequency, since this type of chip should be able to do this quite easily. We set up the BIOS, modify voltages, timings and subtimings, and off we go. For your information, the SA voltage is 1.33 volts, the VDD at 1.59 volts, the VDDQ at 1.47 volts, the IVR at 1.375 volts and finally the MC at 1.425 volts. This is a profile that allows us to compare kits in overclocking.
The good news is that it starts up without any difficulty and our two benchmarks go straight through. We ran several of them in a row, and there were no worries about stability. As you’d expect, throughputs in AIDA64 explode, with read speeds reaching 137410 MB/s, while memory scores in Geekbench3 rise from 13827 points, at 8000 MT/s, to 16370 points, at 8800 MT/s.
Step 3: 8888 MT/s CL36 with 101 MHz BCLK
As my tests progressed, I found it impossible to boot up to 9000 MT/s. The only way to continue increasing the frequency is to increase the BCLK to 101 MHz. This will have the effect of increasing the memory frequency at the same time.
So here we are, with the same voltages/timings at 8888 MT/s. Performance continues to improve, with AIDA64 reading at 138430 MB/s and Geekbench3 memory score at 16556 points.
Step 5: Creating new profiles
At the start of 2024, we’d like to develop our overclocking tests for DDR5 memory kits, whether Hynix A-die or Hynix M-die. We have prepared test profiles that will enable us to evaluate the overclocking capabilities of the kits by checking their stability. We were able to reach a frequency of 9000 MT/s, but not in a stable way, as we’re certainly reaching the limits of our processor.
We’ve also seen that the AMD Ryzen 7 8700G processor seems to have an excellent IMC, enabling higher frequencies with memory kits. We’ll be trying to get our hands on a copy to test this and, why not, include it in our memory overclocking tests.
[nextpage title=”The last word:”]
G.SKill has once again delivered an excellent kit, offering top-of-the-range performance, with this Trident Z5 RGB 8000 MT/s CL40 model. If your processor and motherboard are capable of supporting this XMP, you’ll benefit from very high out-of-the-box performance, provided, of course, you don’t forget to activate the XMP profile in your BIOS.
We can’t do without remembering the performance of the Trident Z5 RGB 7200 MT/s CL36 kit we tested last week. This one, with its 7200 MT/s frequency, is more easily compatible with “mid-range” platforms. By this we mean a non-overclocking motherboard with 4 memory slots and a processor with a decent IMC. As the potential of the G.SKill Trident Z5 RGB 7200 MT/s CL36 kit is very good, you’ll be able to upgrade it to 8000 MT/s CL40, without any worries, if your motherboard allows it. We therefore prefer this kit.
G.SKill Trident Z5 RGB 8000 MT/s CL40 :
We liked:
- Excellent out-of-the-box performance
- The look of the heatsinks, with an aesthetically pleasing RGB, available in white for those who want to match their components
- The quality of finish of the memory modules
- The excellent overclocking potential of Hynix M-die chips
- A voltage of 1.35 volts, which is still very reasonable
- Lifetime warranty
We would have liked :
- A lower price in France than in some German online stores
We hesitated to give it a silver award. And why not gold, when we were so pleased with its performance? Simply because the 7200 MT/s kit from the same series that we tested last week seems more interesting. Indeed, this one is easily compatible with all motherboards, and its overclocking will enable you to reach the performance of the 8000 MT/s kit tested today. Many thanks to G.SKill for making this kit available.