FRONTPAGE Essencore KLEVV CRAS XR5 RGB 32GB DDR5-8000 Memory Kit Review

[Overclockers.com]

Since the premiere of DDR5 memory, Essencore has been playing safe, and we haven't seen KLEVV-branded memory modules at very high frequencies or very tight timings. Recently everything has changed, and the expanded product portfolio now includes memory kits up to 8000MT/s and also lower frequency kits at tight timings. In this review, we will focus on the fastest available KLEVV memory kit, which seems the most interesting for gamers and overclockers.

The tested memory kit is designed for the latest Intel chipsets and supports XMP, so our tests will be performed on the ASUS Maximus Z790 APEX motherboard, which shouldn't limit the hidden potential of the KLEVV CRAS XR5 memory kit.

Click here to read more!

 

[QuantumCookie]

I thought the new DDR5 ram sticks ran off lower voltage.
Says in the article that they can run with 1.55V, is that correct?

Tested Latency (XMP/EXPO)

Profile 1: 8000MT/s CL38-48-48-128 1.55V

 

[Woomack]

Most higher-frequency DDR5s (6400+) are designed to run at 1.35-1.45V. However, as you can see on various motherboard QVL, the top frequencies are sometimes at 1.55V. I'm not sure why this memory kit has such a high voltage programmed in XMP as it's using Hynix A-die IC, and most DDR5-8000 kits with this IC available in stores have 1.45V at the same main timings.
Manufacturers are often setting higher voltages or more relaxed timings than required because of possible weaker IC or weaker CPU/motherboard. However, so far I haven't seen Hynix A-die IC that couldn't make 8000 CL38 at 1.45V. It's usually a matter of the motherboard as not many support DDR5-8000.
As I said in the review, it's not a problem as it's fully stable, and if anyone doesn't feel comfortable, then it's possible to set a lower voltage manually. I tested this kit at 1.45V too, and it was working without issues. The difference in temperatures is in the table, almost at the end of the review.

 

[Hardin]

DDR5-8000! Wow, that is crazy fast now compared to DDR4.

 

[BugFreak]

Simple clean look. I like it.

 

[MaddMutt]

I understand that it does not have a EXPO setting for AMD. Can you still get the memory and then set the timings manually? Is the Rated Speed 8,000 - More limited by the IMC or the Chipset - Using the z690 Apex, Dark, ect, ect..

 

[EarthDog]

Of course you can manually set the speed and timings. You can also set XMP on AMD boards and it should work (so long as the board supports the speeds/memory, etc.). Generally, they are interchangeable, but clearly, the EXPO for AMD is better tested/used.

 

[Woomack]

The only RAM with an EXPO profile above 6400 is the ADATA Lancer (7200 kit was reviewed a while ago). Maybe I missed something from other newer series, but I haven't seen any higher frequency EXPO memory in stores. The main reason is that the support for 6600+ memory kits was added in the latest AGESA, and most memory kits on the market were prepared for sale at least half a year ago. Even the mentioned ADATA kit couldn't be tested on AMD with an EXPO profile because when it was released, there was no AMD motherboard that could even boot at 6600+.

XMP and EXPO profiles in available memory kits are about the same. DDR4 kits sometimes had slightly different profiles, like 1-2 timings were different, but I haven't seen it in DDR5.
All leading motherboard brands provide full support for XMP profiles on AMD motherboards, so it doesn't really matter, but of course, if there is an EXPO profile, then we know that the RAM manufacturer tested it on AMD motherboards.

The reviewed KLEVV memory kit could run at XMP/8000 on ASUS and ASRock X670/B650 motherboards with the latest AGESA b/c. Actually, I couldn't run any kit stable at more than 8000 on AMD motherboards. It was like 8000 is stable, and 8200 barely boots or can't boot at all. Maybe I need a different motherboard or CPU with a better IMC.

 

[Woomack]

Looks like my Ryzen 9 7950X doesn't like DDR5-8000. I could run stability test on Ryzen 5 7600 a while ago with the same RAM, but on 7950X it's crashing after about 15-20 mins. I will play some more with settings, maybe I missed something.
Here is an expected bandwidth/latency and used settings. These settings are passing all benchmarks and game tests, just not extended stability test. Maybe someone find it helpful.



Either way, here is a 1h stability test at 7800 and tighter subs, 1.25V SOC, 1.40V VDD/VDDQ. It doesn't matter much if I set CL34, 36, 38 or 40, bandwidth/latency are almost not changing.

 

[Sologuy]

Looks like my Ryzen 9 7950X doesn't like DDR5-8000. I could run stability test on Ryzen 5 7600 a while ago with the same RAM, but on 7950X it's crashing after about 15-20 mins. I will play some more with settings, maybe I missed something.
Here is an expected bandwidth/latency and used settings. These settings are passing all benchmarks and game tests, just not extended stability test. Maybe someone find it helpful.

View attachment 363041

Either way, here is a 1h stability test at 7800 and tighter subs, 1.25V SOC, 1.40V VDD/VDDQ. It doesn't matter much if I set CL34, 36, 38 or 40, bandwidth/latency are almost not changing.

View attachment 363042

Interesting results there Woomack, also curious how the R5 7600 can handle stability tests yet its bigger brother the 7950X doesn't. They are great looking ram, not that I care that much as long as a kit can fit under my tower cooler & doesn't have ARGB bling. Be interesting what SPD temps while system is in a typical closed up case pan out & at those voltages. The pictures in the review don't indicated if the case they are in is a closed up case.

 

[Woomack]

Interesting results there Woomack, also curious how the R5 7600 can handle stability tests yet its bigger brother the 7950X doesn't. They are great looking ram, not that I care that much as long as a kit can fit under my tower cooler & doesn't have ARGB bling. Be interesting what SPD temps while system is in a typical closed up case pan out & at those voltages. The pictures in the review don't indicated if the case they are in is a closed up case.

I'm not sure if it's a matter of BIOS or some other things. Maybe it's just weaker IMC. 7800 CL34/36/38 (depends on VDD/VDDQ) is stable. Since it's only an additional test, then I won't spend more time checking all the components. It's still good to know that the KLEVV 8000, even though not designed for AMD, works pretty well on AM5 motherboards. Soon there will be two, or maybe three, more RAM reviews.

The test rig is in the ASUS GT502 case. Since there are 9x120mm fans (2x 360mm rads + 3 fans on the side/right), it doesn't matter if it's open or closed. I was comparing results a couple of times, and were about the same for CPU, GPU, and RAM.

 

[Sologuy]

I'm not sure if it's a matter of BIOS or some other things. Maybe it's just weaker IMC. 7800 CL34/36/38 (depends on VDD/VDDQ) is stable. Since it's only an additional test, then I won't spend more time checking all the components. It's still good to know that the KLEVV 8000, even though not designed for AMD, works pretty well on AM5 motherboards. Soon there will be two, or maybe three, more RAM reviews.

The test rig is in the ASUS GT502 case. Since there are 9x120mm fans (2x 360mm rads + 3 fans on the side/right), it doesn't matter if it's open or closed. I was comparing results a couple of times, and were about the same for CPU, GPU, and RAM.

Thanks, I was just raising the point of closed case testing because typically that's what most end users do. So then the test results have some relevance to everyday computing for them.
Looking forward to more RAM reviews.