i7 8700k 4,7Ghz MAX Turbo Limit
i7 8700k 4,7Ghz MAX Turbo Limit
The AC/DC load lines are currently at 0.01. Keeping them on Auto was triggering VR Thermal Alert and throttling, likely because the motherboard was assigning an excessively high value. I plan to adjust the LLC tomorrow after returning from work. If that doesn’t resolve the issue, I’ll attempt to raise the Vcore. Meanwhile, I increased VCCSA to 1.3 and VCCIO to 1.25 to make my G.Skill 3200Mhz memory function with XMP enabled. I gradually adjusted the values from 1.1 and 1.15 until I achieved stability in Prime95 (512k-4096K). The settings are high, but I couldn’t maintain stability with anything lower. The DRAM voltage is set to 1.35 via XMP. I tested 1.36 with VCCIO and VCCSA at 1.15 and 1.2, but it failed in Prime95. It has been running for two hours without errors now, though I’m concerned the 1.3V VCCSA might be slightly too high.
It’s a bit high for a 3200Mhz chip, but if I’m not wrong, higher SA might be necessary to stabilize a weak IMC. A higher LLC could increase IO and SA on my board. You may need to adjust accordingly.
Are there other factors influencing the RAM's stability, and is increasing VCCSA addressing it? I haven't thought of anything, but perhaps I overlooked something.
nightmare392 :
Is there another factor influencing the RAM stability, and is raising VCCSA helping? I haven't thought of anything else, but perhaps I'm overlooking something.
Too high SA can lead to instability—it's one of the most challenging voltages I've encountered.
The guide suggests it's on the higher end but manageable. http://edgeup.asus.com/2017/kaby-lake-ov...g-guide/3/
I noticed the multiplier decreasing to 46. With Intel Extreme Tuning Utility installed, it was set to 46 at some point. Now it drops to 4.7, though I have other issues. I adjusted the BIOS voltage to 1.2V and observed 1152 during full load in Prime95. Raising the LLC to 6 triggered a VR Thermal Alert, suggesting the VRM is overheating. Is it normal for such high temperatures at just 1.2V Vcore? Could there be an error in my settings? I also attempted 1,225V with LLC at 5 and got 1,120V during load.
Nightmare392 :
I have an intel i7 8700K installed on an ASUS Z370-E Gaming motherboard with a Cooler Master MasterLiquid 240 Pro AIO cooler. I attempted to run 4.7Ghz across all cores. I configured the VCORE to Adaptive mode at 1.2 V, and the Load Line Calibration was set to 4 (the calibration failed in Prime95 within 10 seconds). I pushed the power and current limits to their maximum. When I ran the Prime95 Small FFT test, the initial speed was 4.7Ghz, but it gradually dropped to 4.6Ghz and stayed there, with stable operation. In HWiNFO I noticed "IA: Max Turbo Limit" enabled. No other events were recorded—no throttling due to temperature or power draw. The CPU temperature never exceeded 79 degrees during the test. What might be causing the throttling down to 4.6Ghz? I attached a screenshot from HWiNFO.
nightmare392 :
I discovered the issue with the multiplier falling to 46. I had Intel Extreme Tuning Utility installed, and it was set to 46 at some point. Now it drops to 4.7, but I have other concerns. I configured the VCore to 1.2V in the BIOS, which resulted in 1152 during full load in Prime95. If I increased the LLC setting to 6, I received a VR Thermal Alert (the VRM is likely too hot). Should it be overheating so much at just 1.2V Vcore? Why is there such a significant voltage drop? Am I doing something incorrectly? I also tried setting the VCore to 1.225V, which gave me 1120V during load with LLC set to 5.
Try improving airflow over the VRM.
Are the VRM heat sinks securely fastened?
If those don’t help, I’d consider returning or sending a return label.
I adjusted the VRM heatsink screws slightly and received a thermal alert afterward. It appears that setting both AC/DC load lines to 0.01 results in a significant discrepancy between the target voltage and the actual one. I returned to the original settings, but with load lines set to auto, I now measure around 83-84 degrees at 4.3Ghz. Something seems off. I shouldn't be seeing these temperatures with an 240mm AIO cooler.
I adjusted the VRM heatsink screws slightly and received a thermal warning afterward. It appears that setting both AC and DC load lines to 0.01 results in a significant gap between the target voltage and the actual one. I went back to the original settings, but with load lines set to auto, I’m seeing temperatures around 83-84°C at 4.3Ghz. Something seems off. I shouldn’t be experiencing these high temps with an 240mm AIO cooler.
Temperature plays a big role in CPU performance, and the closer the CPU temperature gets to higher numbers, the more it depends on how aggressively the CPU throttles itself. These protective measures are initiated by the CPU and sent to the motherboard to increase fan speed, but if the response isn’t quick enough, internal safeguards kick in. These safeguards might not always show as obvious changes like a voltage drop, but they become clear during benchmark tests when your previous scores suddenly drop.
That said, I’d love to assist you in pushing overclock further. With the cooling setup you’re using, you should be able to hit around 4.7GHz across all cores. I’m running an 8700K with all cores overclocked to 5GHz at 1.325V. To reach that level, I’m currently using chilled water cooling and the CPU has been throttled back. Before that, I was pushing 5GHz before it throttled down, which allowed me to lower the Vcore and achieve better temps.
Are you open to manually adjusting the Vcore? Also, how many LLC levels does your motherboard BIOS support? I’m using an ASUS ROG Maximus X Hero, and we’d like to compare some BIOS details with your ASUS Z370-E Gaming model.
Thank you for your response. I understand that temperature plays a role in CPU performance, but I'm experiencing elevated temperatures with the AIO cooler at standard speeds (MasterLiquid 240 Pro). The BIOS version I have is 0606, and although my BIOS supports up to 7 LLC levels, I'm reaching a maximum of 5 before the VRM starts throttling. This seems inconsistent. At stock speed (4.3 GHz on all cores), I'm drawing around 141W, which feels a bit excessive. When using Auto Vcore, I measure 1.232V at 4.3GHz during a Prime95 small FFT test. I've attached a screenshot from HWinfo showing the temperatures and power consumption. I plan to remove the cooler and CPU, check for bent pins in the socket, and reinstall everything with fresh thermal paste (Gelid GC-Extreme). I'm hoping this doesn't indicate a faulty CPU drawing too much current. Thanks for any guidance you can provide.
The red flag you noticed in your last HWiNFO appears to be linked to your CPU Package Power 141.8w and the clock speed of a 95w TDP CPU, which is significantly higher than expected for a 5ghz OC. It seems there might be an issue with your motherboard.