Need help with OC - single cores on 5950x not holding boost, crazy fluctuation
Need help with OC - single cores on 5950x not holding boost, crazy fluctuation
I'm working with a 5950X and trying to optimize it using Ryzen Master and some YouTube guides. Initially, I managed PPT, TDC, and EDC smoothly, achieving stable speeds around 4.39GHz with temperatures near 77°C during multi-core Cinebench R23 tests. After switching to single-core tuning, my single-core boost fluctuated between 4.7 and 4.1 GHz, dropping back down to 3.5–4.0 before briefly reaching the top speeds again. I spent more time at lower frequencies than higher ones, and only briefly hit peak performance before it dropped. A viewer in a similar setup reported higher single-core speeds (4.9–5.0) with minimal temperature rise, likely due to better cooling. My system has been running since early 2023, and I've experienced unstable boosts during gaming. Is there a simple solution to stabilize these fluctuations? I want to keep maximizing performance without spending extra on the chip or custom cooling. Also, the CPU surface had uneven heat distribution early on, which affected initial temps despite good results. I’m concerned this might be an RMA issue and would appreciate any guidance or documentation you can provide. Thanks in advance for your help!
The focus on 5950 OC isn’t very strong. Ryzen2 was a less aggressive overclocker compared to Ryzen+ and even less so than Ryzen1, making it uncommon. Ryzen3 might break that pattern, but I haven’t heard about it yet. It seems to fit within my understanding of Ryzen3 overclocking. I’m looking for extra details that could help, especially a specific video that shows the setup and might reveal something overlooked.
He began altering values in a series of five parts, as discussed earlier. Before any changes to the precision boost overdrive, I played around for months and clearly saw it through the FPS counter—boost clocks would spike to 4.7-5.0 but then settle for just a second or less before dropping. It’s been a problem long before I realized it was an issue. I’m not used to CPU boost details, so I had no idea boost clocks needed to stay steady for more than a split second.
Observing the differences... handling more than six cores behaves quite differently. With my 5600X I adjusted the maximum core overclock to understand the CPU's limits. I recorded those numbers and fed them into Advanced PBO, which helped me fine-tune it to run at its peak clock for most tasks. My 5900X doesn't act the same. Adding more power didn't help much, but reducing it did. The CPU adapts to whatever it needs, no matter your settings. For instance, I could cap performance at 130a EDC, which works, but if a program demands heavy processing it will push to 180+ EDC. Boosting PPT isn't the solution for everything. Don't overlook Curve Optimizer—it makes a big difference.
I tried adjusting the PBO settings as shown in the video and felt satisfied with the overall clock tuning. Then I focused on changing the values for a single core, which led to unexpected fluctuations. I planned to reach the Curve Optimizer eventually, but the single-core problem stopped me there.
They alternate performance between the two cores during the test. This happens because the benchmark is testing both together and separately. Your observation about the peaks and drops makes sense given the setup. The highest single-core result you saw was 4.85, which is higher than the 4.625 reported.
Heat and fortune? Just a chance pick. Updated July 27, 2021 by Bombastinator
It's a tricky situation. I'm not sure how it all functions. But I've learned how to take advantage of it.