What exactly does overclocking entail? (long, sorry)
What exactly does overclocking entail? (long, sorry)
Hello everyone, thank you for your time sharing your insights. I know OC is meant to enhance hardware performance—affecting RAM, CPU, GPU, and even monitors. I participated in some adjustments, but I’ve noticed varying explanations. Some say OC isn’t harmful and doesn’t reduce hardware lifespan, while others claim it can be risky due to higher voltage requirements. Certain users mention that even with OC, the hardware’s lifespan might only decrease slightly, like from 10 years to 9.5. This makes me confused.
Putting all this together, I realize that if we avoid increasing voltages, OC isn’t considered valid, and it shouldn’t be affecting the lifespan. If I boost GPU clock and memory without changing voltage, I’m still bound by the same limits. Is this still called overclocking? Does it still harm the hardware’s longevity?
Overall, is it worth it? I’ve experimented with undervolting for weaker games like Rocket League, where lowering the stock voltage from 1150 to 950 works well. However, for more demanding titles such as AC: Odyssey, I can’t go below 1100 without issues. Some say the optimal balance is undervolting and overclocking together for stable performance. Considering that higher voltage usage accelerates wear, it seems safer to avoid it unless you’re confident in stability.
The final question remains: is undervolting safe?
Thank you in advance!
Just to clarify your concerns...
yes
probably
it depends on how you define 'safe'
to grasp the response to question 4, you need to recognize that your hardware loses performance at any voltage, consistently. Therefore, nothing is truly secure unless you disable power and inspect the components (perhaps that's why cases with tempered glass sides are so in demand now). For you, safety might mean keeping your hardware functional until the next decade—or even just a few months ahead.
Just to clarify your concerns...
yes
probably
it hinges on your own standards for what's secure
that defines 'safe' for you
depends on your view of what counts as stable
to begin with
You might want to think about how long you plan to keep your gear
until the next big upgrade, or even sooner if you're upgrading
now.
You could also check out this Reddit thread
It focuses on Zen 2, but the ideas are relevant for all CPUs and GPUs.
EDIT added: And honestly...are you really taking the question seriously? If so, the response feels more about passion than tech. To me it's essentially about modifying cars for racing, drag racing, and rallying. So my take would be: "It's about the thrill." This matters because it shapes how I answer question 4—like a racer who accepts engine failure on the first light, you're okay with it when your system crashes during a render."
I thought things through again, but I still hope it won’t be too quick.
If overclocking only gives a small gain—like just 5fps—I’d question whether the risk is worth it, especially if it might shorten the lifespan without any real benefit. Perhaps pushing above 10 would feel justified since it could significantly improve performance in games that run smoothly at around 40 frames per second.
What I see as safe is the information you shared in the link.
It mentions the voltage needed to prevent damage from overclocking, and if the voltage isn’t sufficient, the system might freeze until restarted. I’d consider it safe as long as I don’t risk insufficient power during a crash and have to restart.
You're thinking carefully about this situation. When will you wish you made different choices? How much time do you think you'll need before upgrading? Can you manage the surprises? As they say, "Don't do the crime unless you can handle the time."
Also, remember the main ideas from that linked article: the damage comes not just from the voltage itself, but from heat and current. Only voltage can be adjusted to lessen both effects while maintaining speed. Keeping things cool is essential, but as components get smaller—especially in terms of xistors—they have less space to release heat, causing them to warm up faster. This means you'll need lower voltages to stay within safe limits, even with the best cooling solutions.
First of all, I’m focusing on the temperatures, because voltage influences them, which is why I choose to undervolt to lower the heat (and that’s why I link undervolting with overclocking in the same discussion). But since I’m not very experienced, I can only adjust the GPU settings in AMD mode, but I don’t have the chance to change the BIOS for CPU undervolting either, so I’m currently stuck at a safe clock speed (3.6GHz boost to 4.2GHz) with a reliable cooler that keeps everything below 70°C during heavy use—never exceeding that temperature before now. (The same applies to the GPU.) When will I wish I made different choices? - Regretfully, when I damage my hardware due to my lack of knowledge here. And how long should it last? - Ideally five years or more, since I live in a country with limited income and can’t easily cope with unexpected issues, so I’m hoping for longevity.
GPU performance has always been closely connected to temperature, particularly with AMD products like Vega and now Navi. Manufacturers focus on stable production margins, which often means slightly increasing voltages to provide flexibility. You might find this video helpful—it offers thorough analysis backing its points. While CPU temperatures are a concern, GPUs also require higher voltages during sudden spikes when handling bursty tasks. This contrasts with how GPUs function in games and computing jobs. This difference highlights that undervolting isn't as effective here, just as overvolting isn't always beneficial either. Modern CPUs, especially those from AMD, use boosting techniques that push voltage limits to maintain stability. Increasing voltage quickly leads to overheating, causing clocks and voltages to rise sooner for cooling. In practice, neither extreme is ideal; algorithms often reduce voltage to near-zero levels when necessary to prevent damage and extend lifespan. This is evident in my 3700X, which frequently operates at C1 (partial sleep with lower voltage) over 40% of the time while running complex FFTs for Folding@Home.
Thank you for taking the time to share this helpful video. I appreciate it.