I5-6600K Vcore AsRockExtreme4
I5-6600K Vcore AsRockExtreme4
Philipew:
Yes, it is activated. It reduces my CPU's clock speed, though the voltage remains unchanged if I switch to manual mode in BIOS. It affects the frequency rather than the speed. Refer to the definitions for clarity.
CPU clock speed, also known as clock rate, is expressed in Hertz—usually gigahertz or GHz. This value indicates how many clock cycles a processor can execute each second. For instance, a CPU with a 1.8 GHz clock rate processes 1.8 billion cycles per second.
The clock rate usually describes the frequency of operation for a chip, such as a central processing unit (CPU) or a multi-core processor core. It serves as a benchmark for assessing performance. It's measured in hertz, the SI unit for frequency.
As explained, the clock rate in Hertz reflects the CPU's operation frequency, not its actual speed. A higher clock rate doesn't always mean greater speed; other factors like memory size, cache capacity, and architecture also play significant roles. Keep improving!
L0stChild :
philipew :
bmcat :
Yes, it's activated, it reduces my CPU speed but not the voltage, which remains constant if I switch it to manual in the BIOS.
It cuts down the CPU frequency, not the speed.
Then... what exactly is CPU speed?
You need to test it. That's why all those benchmarks sit at similar levels, and why results differ so much. You can't directly compare 867 on Cinebench R15 (very good) with 7600 (poor) using PassMark. For a car, it's more about the distance traveled per hour—speed is obvious. It's far from straightforward for a CPU, because we talk about execution speed, not displacement, which is much easier (assuming the conditions are consistent, like flat road and steady wind).
The performance of a CPU is influenced by several factors, such as the type of workload it handles—like arithmetic tasks versus image rendering. For example, CPU tests like Prime95 Blend run differently than Cinebench R15 because they measure execution speed in machine code, not just movement. Some calculations require many "carry" microcode instructions, which take up multiple cycles and can slow things down significantly.
It's similar to assessing how much a car can travel in an hour. Would you say a fast Ferrari does more than a slow tractor? Yes, it's true on the highway, but not in the countryside. This is why a car racer and a farmer would have very different views on what speed means.
Think about it...
The CPU Vcore Voltage changes when idle and rises under load (to lower power since current increases). Why would you prefer a different behavior? This is the typical pattern these components follow.
I regret, but your statement doesn't quite convey meaning, as I understand it the opposite is true.
Regarding your second point, I'm attempting overclocking to achieve a stable temperature and extended CPU lifespan at higher clock speeds than the original specifications. I need more control, not wanting the BIOS to behave unpredictably or risk overheating the CPU to 80°C.
On the other hand, my explanation makes perfect sense, even if it might seem counterintuitive. I’m considering investing a substantial amount—USD1M, essentially easy money—I’m eager to see the results. You’ll need solid knowledge or access to reliable resources (like Google) to find trustworthy information, especially after years of experience.
Check this resource:
http://www.tweaktown.com/guides/7481/twe...ndex6.html
It’s titled "TweakTown's Ultimate Intel Skylake Overclocking" and the chart under "Serial Voltage Identification" illustrates how consistently the CPU Auto VCore (SVID) decreases during load, especially at 4.8 GHz where it drops from 1.325 V to a much lower 1.284 V. They also note the significant reduction in voltage from idle to load, which is evident from the graph.
This advice has been discussed before, but it’s widely available online. See here:
https://pcpartpicker.com/forums/topic/13...g-i5-6600k
The opening comment reads: "That's completely normal. It’s caused by a phenomenon called Vdroop. When your CPU goes under load, it leads to a drop in Vcore."
As I previously explained, using the formula [Power (W) = Voltage (V) × Current (A)]—for example, 72 W = 12 V × 6 A—and noting that current increases under load (which isn’t surprising), the motherboard reduces voltage during load as a protective measure. This results in a measurable decrease, such as from 1.356 V at idle to 1.344 V under load—a drop of 12 mV (0.012 V).
If you aim for stable temperatures and longer CPU life at higher speeds, it’s essential to understand concepts like "CPU Load Line Calibration" (LLC) and how it mitigates Vdroop effects. I’ve successfully run my i5 6600K at 4.6 GHz after manually setting the Vcore to 1.355 V in BIOS. It remained stable for extended periods during Prime95 Blend and Heat tests.
Testing with 16 GB (4×4 GB) HyperX Fury Black C14 at 3200 MHz under 1.30 V, and two GTX 970 cards in SLI with GDDR5 memory at over 4000 MHz (without voltage changes), has proven effective. My system temperatures have been kept at or below 60°C using a laser-guided infrared sensor, all while running Prime95 Blend tests with a budget fan cooler.
Don’t dismiss my statements as nonsensical. Regarding your claim about being unaware, I assure you you’re far from fully informed. You still have a long way to go.
If I recall correctly, about 20 seconds ago I mentioned a search engine—let me remind you to explore it before making such a bold claim. Or better yet, learn how to use it properly before attempting CPU overclocking. Good luck! ;-)