Reviews at 13,700k to 13,900k support a request
Reviews at 13,700k to 13,900k support a request
Hello, your question is clear. Would you like to know if the 13700k/13900k supports core down configuration similar to AM4/AM5? For instance, can you change a 7950x3D chip into a 7600x3D, 7800x3D, or 7900x3D by adjusting the CCDs manually? Regarding the 7800x3D with a 1-1-1-1-1-1-1-1 CCD setup in AMD Overclocking, would reducing the number of cores lower the reported L3 cache size in Task Manager? You mentioned hearing about disabling E cores and how that might affect performance, similar to Intel's binning. If a 13900k can be downgraded to a 13600k with more L3 cache, does that impact results? I’m not looking for full benchmarks, just whether this is realistic or if the cache size remains unaffected. Also, you noted that a Windows setup shows different L3 cache depending on core count, and you’re curious about the i9-13900K with E cores enabled.
You can achieve better results on LGA 1700 by toggling individual cores. The main constraint is that at least one P core must be active. While reported cache sizes stay consistent, the actual performance gaps between these chips aren't as significant as in the X3D lineup, though they still exist technically.
I'm trying really hard not to purchase the 13600k and 13900k units just to test this myself, though I don't care enough to run a thorough set of academic benchmarks to truly assess whether the additional 12MB of L3 cache would make a difference. It's easy to show big differences, but proving that an extra 50% more L3 cache would be beneficial feels more like a hardware-focused question. I'll check if the labs would be interested—I think it's something they'd be keen on, since it seems possible.
L1 and L2 drop during the test. A Cinebench test at 6/8 P/E yields 23K scores, matching expectations for a 13,600K result.
Search for a system with lots of caching. It would be great to observe how the results differ. Intel's approaches might vary significantly from AMD's depending on cache size.
It makes sense if someone ran a test. The results would confirm whether the claim holds true. If a 13600k and 13900k (down-configured) showed similar performance with the same clock speed adjustment, it probably means the L3 cache is off. A higher performance than expected might point to more L3 cache being used. That article could be mistaken about E cores not affecting L3 cache reporting.