How To Install HQPlayer Embedded On Ubuntu

[pavi]

i’ve been following along on a i9-12900k, and am experiencing exactly what @jvvita is chronicling. 

[pavi]

so, based on the very useful ongoing discussion, the following works on a fresh ubuntu 20.04 install on an i9-12900k:

 

sudo apt update

 

sudo apt upgrade

sudo apt install linux-image-lowlatency-hwe-20.04 linux-headers-lowlatency-hwe-20.04

sudo reboot

sudo apt install libnuma-dev

 

sudo reboot

 

sudo apt install gnupg2

 

wget -q -O - https://repo.radeon.com/rocm/rocm.gpg.key | sudo apt-key add -

 

echo 'deb [arch=amd64] https://repo.radeon.com/rocm/apt/debian/ ubuntu main' | sudo tee /etc/apt/sources.list.d/rocm.list

 

sudo sh -c "echo '/opt/rocm-4.5.0/lib' >> /etc/ld.so.conf.d/rocm.conf"

 

sudo ldconfig

 

wget https://www.sonarnerd.net/src/focal/libgmpris_2.2.1-8_amd64.deb

 

sudo dpkg -i libgmpris_2.2.1-8_amd64.deb

 

wget https://www.signalyst.eu/bins/hqplayerd/focal/hqplayerd_4.28.3-107amd_amd64.deb

 

sudo dpkg -i hqplayerd_4.28.3-107amd_amd64.deb

 

sudo apt install -f

 

sudo hqplayerd -s hqplayer hqplayer

 

sudo systemctl enable hqplayerd

 

sudo systemctl start hqplayerd

 

sudo systemctl status hqplayerd

 

 

after that, I used the following to identify and install the gpu driver (Nvidia 3080)

 

sudo apt install ubuntu-drivers-common

 

ubuntu-drivers devices

 

sudo apt install nvidia-utils-470-server

 

[pavi]

3 hours ago, jvvita said:

Confirmed. Work for me too on a clean install (i9-10900k no GPU). Thanks

thanks so much, @jvvita & @Miska, for working together to bring this to clarity — very much appreciated.

[pavi]

made the following changes:

 

/etc/apt/sources.list.d/rocm.list

now contains 

deb [arch=amd64] https://repo.radeon.com/rocm/apt/5.0/ ubuntu main

 

/etc/ld.so.conf.d/rocm.conf

now contains

 

/opt/rocm/lib

 

then installed https://www.signalyst.eu/bins/hqplayerd/focal/hqplayerd_4.29.3-117amd_amd64.deb

 

all seems to now work fine.

[pavi]

On 4/4/2022 at 7:49 AM, Rovo said:

January 4th Miska mentions that by using a realtime kernel you cannot have GPU offload because it breaks Nvidia driver (and the Nvidia driver knows this and refuses to work on realtime kernel). Did you get GPU offload working with the commands above?

yes, but you’ve already sorted it out, all is well!

[pavi]

3 hours ago, Rovo said:

If I may ask you another question, I noticed that for the NVidia GPU driver you used "nvidia-utils-470-server". This is all what is needed to get CUDA offload? No "nvidia-driver-470-server" required?

 

these days the current driver is nvidia-driver-510, so if you install that, you should be fine. don't use anything below 490, is what @Miska recommended some time ago

 

so,

 

$ ubuntu-drivers devices

 

should result in something like

 

vendor   : NVIDIA Corporation
driver   : nvidia-driver-470 - distro non-free recommended
driver   : nvidia-driver-510-server - distro non-free
driver   : nvidia-driver-510 - distro non-free
driver   : nvidia-driver-470-server - distro non-free
driver   : xserver-xorg-video-nouveau - distro free builtin

 

and then, 

 

$ sudo apt install nvidia-driver-510

 

and you should be fine. use 

 

$ watch nvidia-smi

 

to observe the gpu load

[pavi]

54 minutes ago, Miska said:

Just an assumption, is this the case?

i’m not sure — i assumed the -server driver did pull in the unnecessary bits, so i’ve only used the regular driver. i should try it out…

[pavi]

On 4/8/2022 at 9:25 AM, Rovo said:

sudo apt install nvidia-utils-510-server (otherwise nvidia-smi is not available!)