Ubuntu
linux package

  1. Bug #1690085
  2. Comment #747

Comment 747 for bug 1690085

Revision history for this message
In , eric.c.morgan (eric.c.morgan-linux-kernel-bugs) wrote :

(In reply to Michaël Colignon from comment #660)
> (In reply to eric.c.morgan from comment #659)
> > I applied the latest asrock BIOS with new options "amd cbs global c-state
> > control" to disable voltage lowering when idle.
> >
> > Even with all BIOS settings, custom kernels and params, disabling C states,
> > power supplies and so forth I'm done. 2 years of this BS.
> >
> > I picked up a 65 watt Intel i9 9900.
> >
> > Good luck to you all and thanks for all the ideas and help. I hope to
> > revisit AMD Ryzen based linux servers in a few years.
>
> I think you had a faulty hardware. With just the typical idle current it fix
> the trouble.
> That happens.
>
> On the other side, Windows kernel has no problem with default parameters, so
> i think something in Linux is not good.

Likely not bad hardware. My 1700 was RMAd for the segfault issue (another pain point), and has passed all testing. Memory has been tested as well, many times over, with different profiles and settings.

Tiny i9-9900 itx ready for transplant. I feel dirty. https://i.imgur.com/fTvWrWd.png