Broadwell-E max safe vcore?
Broadwell-E max safe vcore?
Hi all,
I was wondering if anyone has information about the safe maximum Vcore for Broadwell-E processors? I saw in the Intel CPU temperature thread that the upper limit for 6th generation was 1.35, but it seemed to apply mainly to non-E range chips.
Right now my setup has a 6850k clocked at 4.5 with 1.44v (actually adaptive is 1.42, but cores can reach up to 1.44). The system runs smoothly, and during IBT testing the package reached a max temperature of 83°C, with individual cores around 77°C. For gaming, the package stays near 60°C and cores drop to about 50°C.
I'm using a custom water cooling loop with a 360XE radiator and a 240GTS fan, just for the CPU and a Titan XP.
Is a 1.44vcore safe, even though everyday max temperatures won't get close to 70°C?
If I were you, I’d reduce the voltage just a bit to stay a little safer. It’s not a strict rule about what’s safe versus unsafe. In theory, any voltage above the standard could affect the device’s life, but the exact impact at certain levels is unclear. Factors like your PC configuration, temperatures, usage patterns, and other settings all matter. For example, with Haswell-E, 1.3V was often cited as a safe threshold. But I don’t think there’s a clear cutoff number. In my experience, a jump from 1.31 to 1.325 didn’t noticeably change how long the system lasted. I started using this approach after buying the X99 platform and the 5820k in August 2014, running it at full speed with manual Vcore settings.
That's quite a high vcore for continuous operation. With your 14nm Broadwell-E chip, the maximum would be lower compared to my 22nm Haswell-E. I'm sticking at 1.325 and won't exceed it. For daily use, I wouldn't go above 1.35 if you're comfortable with the demands. The idea of "safe" settings is still a topic of discussion, but anything over 1.4 could harm the CPU over time. You're using adaptive voltage, though. Have you saved any logs from in-game or regular use to track the voltages? They should be lower than the benchmarks.
That's quite a high vcore for continuous use. With your 14nm Broadwell-E chip, your CPU would likely have a lower maximum compared to my 22nm Haswell-E. I'm sticking to 1.325 and won't exceed it. I wouldn't go above 1.35 if you're willing to push it. The idea of "safe" settings is still up for discussion, but using over 1.4 could harm the CPU over time. I notice you're using adaptive voltage. Have you saved a logfile from gaming or regular use to track voltages? It should be lower than benchmarks.
If I were you, I’d reduce the voltage just a bit to stay a little safer. It’s not a strict rule about what’s safe versus unsafe. In theory, any voltage above the standard could affect the device’s lifespan, but the exact impact at certain levels is unclear. Factors like your PC configuration, temperatures, usage patterns, and time in use all matter. For example, with Haswell-E, 1.3V was often cited as a safe threshold. But I don’t see a clear cutoff point. In my case, I bought an X99 platform with a 5820k at launch in August 2014. I’ve been using 1.325 manual vcore and running at full speed of 4.4Ghz ever since. I’ve always optimized this setup. I first experimented with adaptive voltages, but they caused noticeable spikes even at my adaptive setting of 1.325. That’s actually typical for adaptive modes.
My theoretical ceiling is about 1.44v with adaptive. The real overclock voltage is around 1.42v... I might experiment and find out what Vcore I can reach consistently, which was roughly 1.36 from my memory, which seems quite reasonable. I doubt 100MHz would justify the performance boost. I just hope there are users with 1.4v+ chips that have been running for a while and are still functioning well.