Problem with excessive voltage on the CPU has been addressed.
Problem with excessive voltage on the CPU has been addressed.
Looking for a voltage that’s theoretically doable during overclocking—something high enough to risk damage but not obviously dangerous. Something like 1500 volts might seem extreme, yet it could be a plausible figure in an experiment without clear guidelines. The goal is a value that’s technically feasible but laughably excessive.
I envision someone using excessive force and pressing a 7 instead of a 4, which would generate 1.7V—potentially harmful, though the impact might vary based on the CPU. For a completely unaware person, they might attempt to set 12V, believing the power supply cable delivers that voltage, but I doubt most boards support such high levels.
The requirements vary based on the CPU model. Older chips often needed 5 volts, while modern ones prefer below 2 volts or less depending on load. For newer processors, aim for under 1.35 volts when running continuously; under 1.45 with strong cooling for aggressive overclocking—generally not advised. Running around 1.3 volts is common for overclocks, especially with exotic cooling solutions. Exceeding 2 volts usually damages most CPUs or triggers protection mechanisms.
It really depends on the CPU; the board is just as prone to damage from power issues as the processor itself can fail. Older Pentiums operated at 3.5 volts, but pushing that through today’s CPUs would overload the transistors. Most motherboards won’t exceed a few volts unless they’re built for overclocking—though that might not be accurate anymore. The main issue is heat generated by voltage; even with high voltage, poor cooling can cause traces to melt or damage the chip if thermal protection is disabled.
For Ryzen? ~1.85v? A bit older Intel? 2v... but likely less. I own a dead 3770K with an insulated Z77 OC Formula. Not sure how far it went, but it’s done. Probably under 2v. Back in the AXP era 2v was just starting. I’ve been above 2.3v before, but not on anything new. 10-year-old Intel? I’m comfortable with 1.6v if it’s cold or cooling is possible. My 3770K reached 1.575v, and my X5690 hit 1.6v. My 5900X was at 1.55v manually. Gets pretty hot, so don’t push it CPU-wise—until they stop being tough. I haven’t destroyed any units myself. Killed many boards, some RAM, a few PSUs too. I think someone should give it another shot.
It would be reasonable to say there isn’t a standard modern processor designed for such a high voltage, expecting it to function properly.
Yes, 5V seems unusually high for today's CPUs. The power from the 12V+ rail gets reduced to a usable level before reaching the processor. ATX power units typically provide 12V+, 5V+, and 3.3V. With a CPU drawing around 130A at about 200W, there’s still a significant amount of current flowing through it.
It seems like a reasonable guess. I’d be impressed by 2.5v, especially at 5v. You’re right about how intense things can get with 1.55v—most CPUs from the past five years would struggle. The heat from my 5900X at 1.45v is already excessive, and pushing it higher would be a real challenge. I’d probably rely on Linpack for handling such workloads.
Typically desktop CPUs work well at around 1.45V and lose performance only slightly. At 1.5V the performance drops quickly. Between 1.6V and 1.7V you notice significant loss of speed and a high chance of failure, often used for pushing overclocks. Going above 1.7V greatly increases the risk of immediate shutdown. A few processors can handle up to 1.8V without instant failure, while others cannot. For instance, the Core 2 Quad Q6600 usually tolerates higher voltages.
If you froze it and pushed it around at about 2v, it might still work, though it could be risky. It seems like a lot for today’s processors. I checked a forum entry for a 5900X frozen—it was running around 1.75v—and it probably wouldn’t last long on air.