You just received a 5800X CPU/MB and have some questions.
You just received a 5800X CPU/MB and have some questions.
I’m glad I verified it within the return window... I just pressed “memory try it 3800 18-18-18-38” in the BIOS, assuming it would bring it back. So far, I haven’t adjusted the voltage. It’s an old 3200 CL14 unit from my 8086k system that I’m repurposing for a 9900k setup. That’s surprising I’ll be using it again for a third time... This might be a silly question, but why aren’t they offering higher-speed memory or better XMP profiles? It seems like they’re only testing at low voltages.
Also, since I’m heading to Micro Center, maybe I could find something better. I’ve checked and there’s no stock above 3800 on my wishlist, but there was a 5100MT model that performed better even when decoupled—possibly priced higher. Is spending money on this really worthwhile? I know it excels in benchmarks, but I haven’t seen any real-world game performance data yet.
They produce high-speed kits as well. My are 4266mt/s 1.45v CL19 with two 8gb drives. Running at 8700K works fine. Kits reaching 5000mt/s seem to have unstable timing. You're relying on the memory controller, so it probably isn't worth the investment. I've seen 4600mt/s on Ryzen models, depending on how much you're willing to push for XMP speeds. For a setup targeting 3600mt/s with good performance, I think it's achievable. A 1.45v setting might be sufficient. Try adjusting settings and disable power down mode and Gear down mode if you notice odd CL timings.
I’ll try it on pass 2 @ 3866 CL 18, which is better than my 4000 kit even though it’s designed for higher frequencies. I want to hit 3800 at low CAS but I understand it’s challenging. I’m working with a 1:1 memory:IF setting and adjusting timings. I’ve noticed benchmarks showing performance drops after 3800, but is that due to decoupling or something else? I’m curious if pushing further is realistic even if the memory supports it, or should I focus on minimizing AIDA latency for speed/cas balance?
You tested the model with a specific input and confirmed it started working at the standard voltage. It’s now running and you’re comparing its performance to other models like 3800 CL-18 and 16. Whether it offers better results depends on the actual outputs and comparisons you make.
They exceed 3600mT/s. Begin at the start. Raise frequency gradually until it becomes unstable, then introduce a bit of Dram voltage. A value of 1.5 should work for regular use.