Attractiveness of AMD Ryzen despite its instability

Posted on August 24, 2021

AMD are on the uprise for the past several years. In my eyes, their CPUs are clear winners in terms of performance/price ratio. However, people (especially Linux users) should be aware of the potential stability problems!

The issue is that your brand-new PC can suffer from spontaneous reboots. This can start happening days after you’ve built it or (more likely) months – for no apparent reason.

[    0.808978] mce: [Hardware Error]: Machine check events logged
[    0.808980] mce: [Hardware Error]: CPU 19: Machine Check: 0 Bank 5: bea0000001000108
[    0.808984] mce: [Hardware Error]: TSC 0 ADDR 7f0f4932d109 MISC d012000100000000 SYND 4d000000 IPID 500b000000000
[    0.808988] mce: [Hardware Error]: PROCESSOR 2:a20f10 TIME 1620673385 SOCKET 0 APIC 13 microcode a201009

It is such a nasty problem that you don’t even know where to start looking for the cause. It’s not at all connected with the load: it can happen while idling, in a game or during a heavy computation (CPU or GPU); nor is it affected by the temperature of any component. In this lengthy kernel Bugzilla entry many users suggests various workarounds:

The Gentoo Wiki also addresses this issue. If you are lucky, the system will stabilise and work perfectly for 3 weeks, or even 3 months. You’ll think to yourself - “fantastic, the fix has worked!” and get a spontaneous reboot the very next day.

It’s up to you to decide what is more important: unquestionable stability or more cores that have better performance and draw less power.