benchmark9900k gains significantly with cache overclocking (my findings)
benchmark9900k gains significantly with cache overclocking (my findings)
Certainly. Any form of cache optimization can improve performance. Consider increasing the cache frequency to 100 MHz or more. If stable, try another 100 MHz increment. At 4.7 everything works smoothly, but at 4.8 the system freezes after a few minutes without any error messages.
After completing the build, I’ll check it, but I’m hoping it will last me through high school and beyond. I don’t want to take any chances (though I’m sure it won’t cause any damage). Lol
Voltage, heat, and current damage your hardware—not just the clock speed. Running a CPU at 1.25 volts with 5 GHz across all cores works as long as it drops to 1.25 volts at 3 GHz; I haven’t seen any degradation even under high voltages or heavy load. Just keep things cool and avoid pushing it beyond 100 MHz with extreme voltages. (Except you’ve been in school for 20 years!)
It's not about issues, but a 100MHz core clock is more valuable than an equivalent uncore clock at that stage. At around 5.3GHz, the core clock is likely quicker than the 5.1GHz uncore and even the 5.2GHz version.
Overclocking cache ratio presents additional challenges with 9900k and similar hyperthreaded processors. This is due to the heavy reliance on L3 cache in hyperthreading, where logical cores depend on it for proper execution. Virtualized register storage at L0 becomes crucial for logical threads to function correctly. Errors at L0 are more frequent when hyperthreading is active compared to when it's disabled, often leading to crashes or BSODs. VCCIO may assist in resolving IMC-related L0 issues, while Vcore can be useful in other cases. Testing stability against IMC/L3 problems is recommended using tools like prime95. Some users report success with specific configurations, such as the 112k in-place FFT variant, though adjustments may be needed for consistent performance.
Overclocking cache is quite simple. I've done it on multiple CPUs (4790k, 8600k, 8086k and now 9900k) whenever there were two situations. It performed perfectly—100% stable and completely reliable in every stress test or it would crash and freeze the whole system right after startup. The same went for the 9900k 4.7 and 4.8 settings; no BSOD during boot.
It's challenging since some individuals struggle to surpass x44 cache on their 9900k without significant voltage increases. This is common; just because it didn't occur for you doesn't rule it out. I also can't reproduce your results. Small gains here. 5.1 / 4.7: CB 15: 2273 (all five loops) CB 20: 5409 (looped 3600 seconds, halted after five times the scores due to identical results). 5.1 / 4.8: CB 15: 2285 CB 20: 5411 (identical across all five loops). I didn't apply realtime priority. I only restarted Windows, waited a minute, then ran.
So I ran all the tests once more and added Blender from Steam with the BMW CPU Render Scene. I took a screenshot from the 4.9 / 4.7 results because it was unbelievable. The numbers came from yesterday and today on a brand new Windows 10 Pro setup. I lowered the voltage a bit, but the room stayed warmer. Intel I9 9900K R0 Stepping with a Corsair H115i (280mm) in a single closed loop cooler. The mainboard was ASRock Z390 Extreme4 with all power limits at maximum and Powersavings turned off. Every overclock was solid, no AVX offset, fixed voltage. Aida64 Extreme CPU, cache, FPU & RAM temps and power usage. (Average temp after 10 mins with 22°C outside) Core/freq 4.9 / 4.3 1.264v 184W 77°C 5.0 / 4.7 1.35v 229W 90°C 5.0 / 4.5 1.35v WHEA Error / L0 Cache Error / BSOD Cinebench R15 4.9 Ghz Core 4.3 Ghz Cache = 2073 5 Ghz Core 4.3 Ghz Cache = 2133 4.9 Ghz Core 4.7 Ghz Cache = 2171 Cinebench R20 4.9 Ghz Core 4.3 Ghz Cache = 4911 5 Ghz Core 4.3 Ghz Cache = 5115 4.9 Ghz Core 4.7 Ghz Cache = 5172 Blender benchmark BMW (version 2.80.75) 4.9 Ghz Core 4.3 Ghz Cache = 187 seconds 5 Ghz Core 4.3 Ghz Cache = 175 seconds 4.9 Ghz Core 4.7 Ghz Cache = 153 seconds Timespy CPU benchmark 4.9 Ghz Core 4.3 Ghz Cache = 11984 5 Ghz Core 4.3 Ghz Cache = 12192 4.9 Ghz Core 4.7 Ghz Cache = 12401 Average FPS after 10 minutes or a full race for Project Cars 2 (RTSS) in Novigrad. 5.0 / 4.3 = 138 FPS 4.9 / 4.7 = 147 FPS Assassins Creed Odyssey in Athens. 5.0 / 4.3 = 100 FPS 4.9 / 4.7 = 105 FPS Project Cars 2 Racing 20 cars at Côte d'Azur. 5.0 / 4.3 = 244 FPS 4.9 / 4.7 = 261 FPS ARMA III Start Mission. 5.0 / 4.3 = 99 FPS 4.9 / 4.7 = 112 FPS Fallout IV in Boston. 5.0 / 4.3 = 98 FPS 4.9 / 4.7 = 105 FPS