No, CL/Latency doesn't influence compatibility.
No, CL/Latency doesn't influence compatibility.
I'm setting up a new PC with the Asus B650 Proart Creator + Ryzen 7 7800X3D and the 2x16 G.Skill Flare X5 DDDR5-6000. The motherboard doesn<|pad|>, but it can't detect the RAM sticks. The B650 shows the Flare X5s on its QVL, but with a lower latency (CL36) than what I have (CL32). Latency differences can influence compatibility, but they don't always cause major issues. You might need to get the CL36 version now or look for a workaround.
Each set varies, you received one that hasn't been verified. It remains likely to function, just turn on XMP/EXPO and test. Your unit should be acceptable since it's within spec. If you notice any instability with the RAM module, consider recommending a 6000mhz CL 30 kit on the QVL list for the motherboard.
There are four key factors influencing memory compatibility: the PCB, the memory IC, and the operating frequency. The PCB tends to remain consistent across kits with similar ratings, ensuring uniformity among those. Frequency stability also plays a role, leaving the IC as the primary consideration. Motherboards require tailored support for each memory IC to function correctly, and CL30/32 kits differ from CL36 kits on average. CL36 kits are typically Samsung-based, while CL30/32 kits represent various Hynix variants. Despite this, Hynix memory is currently among the top performers, prompting manufacturers to prioritize it. Consequently, CL32 kits likely offer better compatibility than CL36. The absence of your kit was likely due to incomplete testing, not a lack of functionality.
The issue seems to lie with the board or possibly faulty RAM. Boards are generally the most prone to errors, especially near RAM modules like XMP or mismatched ICs. Most reputable manufacturers don’t support such components, as they’re among the top DDR5 options (though Hynix diesels are limited to around 7600 MHz). A 6000 MHz should not cause problems with the CPU’s memory controller. CL timing is standard, so timing issues are unlikely. RAM manufacturers typically don’t design kits that only work with one platform. Since it’s an Expo board, older BIOS versions might have XMP conflicts, but that shouldn’t be the main concern. It probably points to the board itself unless the RAM is clearly defective.
I just reinserted the RAM three times and after the third try it finally worked, so it might have been the board or possibly something in the slots.
All that remains is to perform a stability test on the RAMs I personally use, such as prime95 largeffts. There are alternatives like tm5 (anta777 profile), HCI Memtest, etc. The gold contacts on the RAM can become dirty and prevent detection. Cleaning with 99% IPA mixed with a tissue or even a wet tissue works well, though I tend to use my fingers which can also cause finger cuts—something that has happened twice already. Apparently, a fiberglass PCB has some sharp edges, and DDR5 RAM completely overlooked that.
It means RAM might function partially, it could fail, or it may lead to problems during regular use.