Some motherboards are optimized for specific Linux distributions, offering better performance and stability.
Some motherboards are optimized for specific Linux distributions, offering better performance and stability.
Hi, let's go over what I've encountered with Linux setups so far. I started by installing Ubuntu 1904 on my main PC, which was running Windows 10 Pro in dual-boot mode. My first challenge came when I tried to use a higher clock speed—4.8GHz at 1.25V for gaming on Windows. Ubuntu didn't accept that setting and kept reporting hardware issues. Eventually, I switched to a 4.5GHz configuration, which Ubuntu considered stable. That forced me to lower my own system to 4.5GHz as well. For comparison, I can run 4.9GHz at 1.20V on Windows. I also experimented with Manjaro, but it froze twice during boot and installation, so I stopped there.
My PC specs are: Intel Core i5-8600K @ 4.9GHz, 1.20V, 16GB Corsair LPX RAM at 3000MHz, Asus Prime Z370-P motherboard, ASUS GTX 1060 with 3GB VRAM.
On the topic of motherboards, do some models really support Linux better than others? I've faced problems on my current system with various distros and wondered if it's the hardware or something else. I tried Ubuntu 1804 live-booted on my brother’s ASRock B450M Pro4 motherboard running a Ryzen 2200G, which worked smoothly. If I get a new board, I’d like one that allows stable overclocking—ideally above 4.8GHz—and supports dual-booting Windows and Manjaro without crashes.
I also explored Manjaro on my laptop (an Asus ZenBook with an i3-8130U and integrated graphics). It ran well, though I had an issue where the Ubuntu 1904 installation got corrupted after a restart.
In short, my PC struggles with high overclocks on Ubuntu, crashes when using Manjaro, and works fine with Manjaro Gnome on my laptop. The i3-8130U and integrated graphics seem to handle Linux better than my current setup.
Based on what I understand, there aren't necessarily boards that perform better with Linux than others, though some might offer unique features not fully supported by the OS. This shouldn't impact the stability of an overclock, which typically works well regardless. It's likely that Linux is more prone to instability overall. Even though desktop versions get regular updates, we might see greater flexibility for higher overclocks in the future.
Yes, mostly due to drivers. Haven't heard much about overclocking being worse on Linux depending on the motherboard but it could conceivably depend on the quality of the chipset drivers - this would require some testing with different motherboards. Unfortunately there is virtually no information on this, no major outlets test overclocking on Linux (to my knowledge) and besides due to the nature of overclocking it would be hard to predict your results anyway. Level1Techs generally test motherboards on Linux (though again, no specific mentions of Linux OC) so if you're looking for a general compatibility assessment you can check them out .
This topic has been discussed often, with a common idea that overclocked CPUs don't pair well with Linux. I'm experimenting with overclocks for Linux distributions and have a decent 8600K setup from Windows feedback. I'm comparing different boards using the same 8600K. Recently, @Twilight suggested Level1Techs to me while searching for a new Z390 board, which sparked the whole motherboard conversation. The testing process is helpful since I've been interested in Aorus boards for a while.
It's not something I'd claim. Linux generally performs better than Windows in terms of efficiency. Greater efficiency reduces the need for intense overclocking to achieve similar results. The best choice depends on your specific needs and intended use.
I intended to rely on Ubuntu as my everyday desktop environment, covering both regular tasks and gaming. Unfortunately, after getting heavily overclocked, I found it challenging to keep my 4.8GHz+ setup running smoothly for Ubuntu. The same issue appeared with Manjaro, where installation failed entirely. I recall using Ubuntu for BOINC when it was around 4.5GHz, and the Pentathlon workload pushed my 4.5GHz 1.25V OC 8600K to a sweltering 70°C after two hours of continuous use across all six cores. This was significantly hotter than what I’d experienced at 4.9GHz with a lower voltage. In short, I’m focused on preserving the overclocks I invested in. If I didn’t need such performance, I would have chosen a simpler processor like a Ryzen 1000.
based on what I've seen, it's usually not about overclocking. I've noticed my brother's system would crash when Windows started but stay stable with Linux, and hardware issues often caused problems. That makes it more sensitive. My main PC ran at 3.5ghz and was just as shaky in Linux as in Windows—sometimes freezing every 12 hours, other times crashing once a day. I've followed Wendell for years; he's really skilled. This is mainly due to the scheduler. For simple tasks like gaming or browsing, it probably doesn't matter much. It might affect low-end setups more because fewer programs run at once, but on a 8600k at 4.8ghz it should handle decently. Do you really feel a difference in performance? In games, the 1060 would definitely slow things down.
I conducted some tests on Windows and Linux using BOINC. Performance tends to be better on Linux at 4.5GHz compared to Windows at 4.8GHz. That’s the usual outcome. As a regular user, this is often the case. Certain functionalities perform more smoothly on one system versus another. You might consider dual-booting Ubuntu and Windows (if you haven’t already) and adjust BIOS profiles depending on the apps you need.