Nicely done! 👍🏼 Do you think the lack of increase in gaming performance could be because the memory controller isn't fast enough to feed the cores? If so, X3D versions might scale even better this generation.
it could be that the 2080ti (2 generations old) he's using is maxed out by the cpu even at ambient temps, so raising cpu speed wont realy help. If he had a 4090, the CPU would be a bottleneck, so higher CPU speed would show improvements.
The lack of gaming performance scaling with much higher frequency is a bit puzzling, however there is an element of GPU-boundedness with this RTX 2080TI. That said, usually CPU frequency always helps performance even if it's just a couple FPS. Even in extreme GPU-bound scenarios (because the CPU is always there in terms of fetching data from memory or processing some data). For example, I've gotten to 207 FPS in Tomb Raider with this GPU and another CPU. Also Final Fantasy sometimes goes over 200FPS with other CPUs. It feels weird. I also tried the RTX 4090 with this CPU on ambient and there was a similar lack of scaling. I don't think it's related to memory. Otherwise we'd see performance improvements with FCLK/UCLK/MCLK increases or memory timing adjustments. My gut feeling says the bottleneck is somewhere in the uarch backend getting the right data to the right places at the right time. Because it doesn't matter how fast you can process data if you don't have the right data to process since you'll need to wait additional cycles. This is just speculation, however.
Might be interesting to try a tec system set up to hold the cold plate just above ambient to get optimum ambient temperature performance. For gaming loads this wouldnt be too much for power probably.
@@Sir_Defyable Most get that, but his point was in comparison to what others have achieved with the same cpu on LN-2. SB is right on the money though, pretty impressive just using PBO and E-Clock!
