Issues with voltage readings on this processor model
Issues with voltage readings on this processor model
Hello. I recently installed a new build and reviewed various hardware monitoring tools, which revealed inconsistent CPU voltage readings across different programs. The data from HWMonitor shows significant differences (1.435 vs 1.214). CPU-Z matches HWMonitor but comes from the same manufacturer, while Intel XTU displays another value. All these monitors operated separately. According to the i7-13700K specifications, a safe voltage is up to 1.6 V, though typical overclocking settings often reach 1.2–1.3 V. I understand these tools aren’t perfectly precise, but the higher readings worry me, especially since HWMonitor might be correct. I noted that after undervolting by 0.115, the display changed from above 1.5 V to below it. Is there a way to resolve this, or is this just how these monitoring programs behave? Please let me know if you can help. Thanks.
I’m a bit wary of voltages exceeding 1.5V, especially for daily use. For testing I can push up to 1.6V, but I believe the safe upper limit comes from the CPU’s capabilities rather than Intel’s guidelines. The advice to stick between 1.2 and 1.3V during overclocking stems from cooling challenges beyond that point. My 13700K, for example, stays stable under heavy load (Prime95 Small FFTs) at around 1.24V die temperature with a custom loop running on P cores, E cores, and Ring. Different voltage readings often reflect various sensors—typically three or more for the VCore depending on the board design. HWMonitor and CPU-Z usually read from the SuperIO sensor, which can be 30-50mV higher than the actual core voltage due to socket impedance. Unless the reading is from a different location, it’s likely just a minor offset. As long as temperatures are normal, the voltage values should be reliable.
Use HWiNFO for core voltage readings. The VID value shown here isn't the real voltage reaching the CPU. You can safely disregard it.
The variation is much greater than 30-50 mV. I also believe anything over 1.5 V is worrisome, which is why I’m trying to understand this. As noted before, it was well above 1.5 V before I reduced the voltage. If any sensor reads above 1.4 V, that would be quite concerning; I don’t think the impedance would create a 200 mV difference compared to the hardware’s core voltage on HWINFO, would you?
200mV is quite significant, though not impossible. A gap of 200mV between the VRM output and the CPU isn't unusual, especially if the HWM sensor is accurate. It's also possible the voltage sensor itself is faulty, causing the discrepancy. To determine the most precise sensor, you'd need to measure directly from the back of the socket or check the VCore capacitors on a higher-end board. If you must rely on one reading, go with that one—it’s likely the most reliable.