Comparing the I7 990x and i9 10980xe highlights their differences in performance and features.
Comparing the I7 990x and i9 10980xe highlights their differences in performance and features.
I own an i7 Extreme 990 with a base clock of 3.47GHz. I'm considering upgrading to an i9 10980xe. I understand the clock speed will drop compared to the 990, but I'm not planning to overclock unless it's safe and beneficial. How significant would the performance drop be between the i9 Extreme and the i7 Extreme? Since I'm not into gaming, I'll focus on the potential benefits. Intel is the only option right now, so let's keep the comparison clear.
Nope, I actually mean the i7 990x: https://ark.intel.com/content/www/us/en/...l-qpi.html
So you have no idea on IPC difference (how much work a single clock cycle can do, often tested by locking multiple CPUs at the same frequency), no idea on core count and no idea on alternatives in the market (even within Intel itself), instead thinking "clock speed rules". That's a lot of extra reading to do. 1. IPC, Cascade Lake X offers at least 33% gain here in older workloads like Cinebench R15. I cant find data with such old CPUs on new benchmarks but the lack of new instruction set support on Gulftown (architecture of 990x) like AVX (of any variant) or TSX means it will be at least a 3 digit percentage difference in new workloads that do use new instruction sets. 2. Core count. 990x has only 6 cores, 10980XE has 18. HTF is this in the same realm? 3. Alternatives. 109xx series on LGA2066 are in really short supply, even worse than Threadrippers in availability (regardless of country) unless you consider review units. Previous gen 9980XE will be the way to go if you need 18 cores. 4. Alternatives if 6 cores 12 threads are enough and you just want faster cores: incoming i5 10600k/kf (k means unlocked, f means no iGPU) (not available yet) has the same 6c12t setup, all the IPC gains and around 4.6-4.7 all core frequency. 9700F is locked, offers 8c8t setup and is currently available but the socket will be replaced for 10th gen so the only upgrades will be the 8c16t i9 on LGA1151.
I understand the basics about thread count, but from what I see, the BC for the i9 10980xe is actually lower than the i7, which makes sense since it should run a bit slower.
I own a Xeon similar to the 990X (X5675, boosted to 4.5GHz — surpassing your 990X's capabilities). I also possess a Ryzen 3600 that remains at its base speed. Both models feature six cores and twelve threads. In benchmark tests, the Ryzen (~4.2GHz) outperforms the Xeon (~4.5GHz) by roughly thirty-five percent. My Xeon, running at stock speeds (~3GHz), lags behind the Ryzen by about half. While base frequencies matter, it's the actual performance under load that truly counts. During recent tests, Cinebench R15 showed my Xeon achieving 140-145 points at overclock and 95-100 at stock, whereas my Ryzen delivered 190-200 points. One test not reliant on newer instructions helped highlight the difference in single-core speed. It's clear that raw GHz isn't the sole deciding factor; IPC (instructions per cycle) is more telling. For modern processors, turbo speeds often matter more than base rates. As for the 10980XE, it's a solid choice for targeted configurations, but alternatives exist across different price ranges depending on your intended use.