Seeking advice on advanced 7950X3D overclocking past 5.9Ghz on single or dual cores
Seeking advice on advanced 7950X3D overclocking past 5.9Ghz on single or dual cores
The displayed results indicate solid performance with decent cooling, achieving +200Mhz boost via Curve Optimizer and Boost Clock Override. I have ample thermal margin, so I'm focusing on fine-tuning per-core power usage and boost clock adjustments to optimize each core's operation. Should I prioritize pushing the stronger cores further or balance the distribution? This is where I'm currently positioned, especially since I'm still figuring out the best approach for different CCDs.
These readings reflect performance under stress or benchmarks like Cinebench and AIDA. Typically, top cores handle less aggressive voltage reduction since they aim for maximum speed before power drops, while lower-end parts may run fine even at very low temperatures but won’t reach peak performance. It’s a trade-off, and the 3D VCache feature helps improve efficiency by caching results.
I used HWiNFO for a couple of hours, doing some benchmarks like Cinebench R23 and AIDA64's SHA3. I also did my usual tasks. For the benchmark image, I focused on peak performance and temperatures during typical usage. For cores #8 to #15 (CCD2), you should shift from -20 to 0 as clocks stay fast, and for cores #0 to #7 (CCD1), aim around 5.25 GHz, as that’s the typical limit there.
I've also examined the behavior of Core VIDs in Cinebench R23. This might be helpful.