I don’t think it’s the scheduler this time with a single CCD, but there is significant difference. These tests focus on compute and productivity with almost no games, so most of the difference could come from this bias. Another possible option is the power profile (EPP balance_performance) holding back the 7700x on linux.
The draft is pretty good. Only a few points to consider changing:
That’s more than enough. You can’t do any more.
As an Android flavour it should be safe after uninstalling all apps associated with the university. Did any of them need a “device owner” permission? That’s the only way to be more persistent on Android without root access.
The mobile and TV clients are often limited to the codecs with hardware acceleration. Or just selecting a lower bitrate on the client will cause transcoding.
I think calling it a “cache” is not precise. The primary function of the DRAM is to hold the dictionary for translating logical addresses (e.g. sectors) from the OS to the physical addresses (which NAND chip, which bank etc.). This indirection is needed for the controller to do wear leveling without corrupting the filesystem.
On a SATA SSD without DRAM each read IO could mean 2 actual reads: first the dictionary to find the data and than the actual data being read. As you said HBM helps by eliminating this extra read.
The read and write caching is just a use of the remaining DRAM capacity. Since modern Operating Systems use the general RAM for the same function it is usually just a small increase to the throughput.
There is an even more relevant video of using external storage trough USB. He recommends using software raid:
Are both drives fully encrypted with LUKS? Is trim enabled in both crypttab and fstab?
Thanks for the links! I updated my config from z3fold to zsmalloc and adjusted the vm.page-cluster to test these out.
Reading a bit more, I think when using large max_pool_percent (>30) with Zswap the two solutions are more similar than not. A crucial difference is what use-case is more acceptable since Zswap can cause unresponsiveness (and potential lockup) under high memory pressure. While Zram could result in an OOM crash in a similar worst-case scenario.
Btrfs with compression enabled and subvolumes set.
And enable/automate maintenance services for BTRFS. For example: balace should be run on heavily used system disks or scrub could help detect errors even on single disks.
ZRAM (With proper sysctl.conf like PopOS does).
Could you explain the preference of ZRAM over ZSWAP? I thought the latter was the more advanced and better performing solution. Is there some magic in Pop’s config?
Happy to help! Tough you are right, this is a rather generic error that doesn’t help much just confirms that the GPU is the issue.
At this point it could be a driver issue since there are similar open bug reports. A hardware problem is still possible since you previously said that it’s unstable on windows too, and power related issues can also lead to this error message.
Most distros use systemd and its logging solution: journald. You can use journalctl to read the logs around the time of the crash for e.g.:
journalctl -S -5m this shows the last 5 minutes. Use this when a game crashes but the system continues working and did not reboot.journalctl -b -1 -S -10m this shows the last 10 minutes from the previous boot. Use this if the crash froze the whole system and rebooted.Look for red lines (errors) and what wrote them. AMD GPU faults usually have the ‘amdgpu’ mentioned, memory errors could appear as ‘protection fault’.
Did you check the system logs to see what caused it?
Many things can result in seemingliy random crashes. Any overclock (including XMP and Expo) or undervolt or even a bios version can be problematic.
I would check first if it’s stable on windows.
What filesystem are you using? Is it encrypted?
Could you run a benchmark to verify if reads and writes are both affected? KDiskMark is like crystaldiskmark or Gnome Disks has a built in benchmark.
Many are already mentioned but there is a lack of SoL and KyoAni shows so these are what I missed:
Filesystem permissions
For many apps it is not an issue and provides additional security but in other cases it’s very annoying and not trivial to fix.
Example1: opening a .docx from Thunderbird flatpak with OnlyOffice flatpak does not work out of the box.
Example2: mpv and VLC flatpaks work well for local files, but fail to open network shares from Dolphin.
I think a possible solution would be runtime permission dialogs when denied access.
I’m not 100% sure, but for me it caused a similar “freezing” or unresponsive experience when the daily cleanups run in the morning. If there was a freeze after every (even short) sleep and resume that might be a different issue.
Yeah it’s disappointing that such an annoying bug is still present and quotas are enabled without warnings. You could continue using Timeshift, the only feature quotas provide is the individual disk usage of the snapshots.
Anyone looking for the solution: https://www.suse.com/support/kb/doc/?id=000020696
What did you use to manage snapshots? For e.g. timeshift enables quotas and that can cause freezes when deleting old snapshots.
That ATX board would be great. The mATX B650M PG is also better than the previous one, it is good enough. If you can find the B650M-HDV/M.2 in stock that is even better if you don’t need 3 m.2 slots.
That monitor was indeed a lucky deal. It looks to be a good combination for this setup.