As a PC enthusiast, you probably get the itch to tweak or optimize your PC every once in a while. Whether it’s about higher clock speeds, lower temperatures, or more FPS, the work is never really over. In this race to outperform yourself, you might inadvertently find yourself believing several myths or lies about PC hardware tweaking that are only holding you back from achieving a more performant, stable, and cooler machine.
Many of these lies might just be wishful thinking that you cling to because it’s easier to stick to the familiar than venture into the unknown. Let’s address 5 such myths surrounding overclocking, undervolting, CPU cooling, and motherboard VRMs.
Related
4 PC performance myths that hold no water
Don’t let these myths keep your PC from performing its best.
5
Passing a stress test means a stable overclock
If only it were that easy
Overclocking the CPU might not result in the same gains it used to years ago, but that doesn’t stop many enthusiasts from doing it. The sense of achievement in eking out a few percentage points of extra performance can compel many users to sink hours into getting a stable overclock. Tell me if I’m describing you here — you use Cinebench or Prime95 to run a stress test for each overclock, and congratulate yourself on a successful test, thinking your overclock is now “stable”.
Contrary to what many users think, a few synthetic CPU tests can’t determine the stability of a CPU overclock. Even if your CPU doesn’t crash during a Cinebench run or strenuous Prime95 pass, it can easily crash during gaming or a sufficiently long video render. Real-world scenarios can expose the overreaching effects of a CPU overclock on the rest of the system. Your PC might crash due to memory instabilities or other cascading effects of the overclock on the machine. Only after sustained real-world usage without encountering any instabilities can you confidently claim to have overclocked your CPU.
Similarly, to test your GPU overclock, play a few long sessions of games even if your GPU doesn’t crash during a FurMark, Superposition, or 3dMark tests.
Related
Top 9 programs to stress test and benchmark your Windows PC
You should stress test your PC to make sure everything is running smoothly
4
Custom loops are at high risk of leaking
A loop is only as strong as its weakest link
Over the years, custom watercooling has lost favor even with enthusiasts, since AIO liquid coolers and air coolers offer almost the same cooling performance, and the hassle involved in creating your own custom loop is worth it only if you need the clean aesthetics and superior silence that it provides. However, the reason not to bother with custom loops that many PC builders give themselves is that they’re highly prone to leaking. This is a bit ingenuous since leaks can also occur in AIO coolers, and all that really matters is the quality of the construction.
If you have the know-how to craft a robust loop with the right fittings, and are confident in your abilities, then there’s no reason a custom loop would be at any higher risk of leaking than an AIO liquid cooler. Custom loops might need more maintenance, yes, but the risk of leaks is often overblown. Even if you aren’t skilled in constructing your own custom loop, you can pay a vendor to do it for you. There’s no reason to shy away from custom watercooling if you’re really invested in it.
3
CPU delidding is the best way to lower temps
Most impractical, maybe
CPU delidding is perhaps the craziest thing you can do with your CPU — removing the lid or the IHS from the CPU so that you can cool it better is the most “enthusiast” thing you can do with your PC. However, is it fair to term delidding as the be-all and end-all of CPU cooling? Sure, mounting the CPU cooler directly on the CPU die has its benefits, but how universal is delidding? Older processors had poor-quality thermal paste between the die and the IHS, so removing the IHS and using a better thermal paste or liquid metal could lower CPU temps by up to 20℃.
Intel and AMD CPUs manufactured in the last 7 to 8 years, however, started using a higher quality TIM (thermal interface material) between the CPU die and IHS, so delidding stopped delivering huge reductions in CPU temps. Besides risking damage to the CPU and voiding the warranty, CPU delidding is worth it only if you’re using liquid metal as the TIM between the die and IHS. Since that further shrinks the number of users comfortable with the process, delidding has become little more than a vestige of an old era.
Unless you have one of the flagship CPUs on the market, and are planning to break overclocking records, delidding your CPU is simply not recommended. You can easily achieve great CPU temps with AIO liquid coolers or custom watercooling, without damaging your CPU or shorting your motherboard with liquid metal.
Thermal Grizzly Intel 13th Gen Delid-Die-Mate
Thermal Grizzly’s Delid-Die-Mate is a delidding tool by Der8auer, ensuring simplified IHS removal from Intel’s 12th and 13th Gen CPUs.
Related
Thermal paste vs liquid metal for your CPU: Which is better?
Liquid metal is the new thing and promises greater cooling. But is it really worth the risks?
2
More VRM phases mean a better motherboard
Phase this one out ASAP
Motherboard manufacturers love to market 12, 18, and even 24-phase VRM designs to overclocking enthusiasts, creating the illusion that the number of VRM phases alone decides the overclocking capabilities of the motherboard. While it’s true that the more VRM phases, the more efficient power delivery can be to the CPU, many other factors play a role in making a motherboard ideal for overclocking.
First, the components of the motherboard, such as capacitors, chokes, and MOSFETs should be of high quality to withstand long-term overclocking. Second, a durable and well-designed PCB is also required for efficient power delivery to the CPU. And third, the heatsinks on the VRM should be able to dissipate heat adequately during overclocking. All these features together determine whether your motherboard is “overclocking ready”.
Many manufacturers even advertise fake VRM phases by doubling the phases electronically without adding additional power delivery circuits. Hence, it’s always necessary to dive into the technical details of a motherboard, and analyze reviews (if available) before trusting a motherboard based solely on the number of VRM phases.
1
Undervolting hurts performance
Not if you do it right
I’m sure many PC users have constantly come across the benefits of undervolting — better thermals, lower noise levels, and higher power efficiency — but ignored it for fear of lowering their CPU or GPU performance. While undervolting involves lowering the voltage, and hence, wattage of your CPU/GPU, it isn’t necessarily accompanied by a drop in performance. In fact, the reason most people undervolt their hardware is to enjoy the same clock speeds at a lower voltage.
Undervolting involves utilizing the voltage-frequency curve of the CPU or GPU to dial in a lower voltage setting while maintaining the same frequency. This results in lower power consumption, lower temperatures, and a longer lifespan for the hardware without sacrificing performance. Programs like MSI Afterburner and Ryzen Master have built-in features for undervolting the GPU and CPU respectively. Intel CPU owners can undervolt their CPU in the BIOS.
Related
Overclocking vs undervolting: Which one do you prefer?
Are you a fan of overclocking, or do you prefer the superior thermals provided by a stable undervolt?
Stay ahead of persisting PC hardware myths
Myths and misconceptions about PC hardware, especially surrounding overclocking, undervolting, and cooling, can persist for years. Each one of us is bound to believe in one or the other hardware tweaking lie, but the important thing is to stay open to new information. Once you encounter evidence disproving a lie you believed in, be ready to shed it and move on.
#lies #tweaking #hardware
source: https://www.xda-developers.com/pc-hardware-tweaking-lies/
