diff --git a/manuscript/images/cryptominer_monitor.png b/manuscript/images/cryptominer_monitor.png new file mode 100644 index 0000000..af71d1a Binary files /dev/null and b/manuscript/images/cryptominer_monitor.png differ diff --git a/manuscript/images/donate_bitcoin_qr.png b/manuscript/images/donate_bitcoin_qr.png deleted file mode 100644 index 39105bb..0000000 Binary files a/manuscript/images/donate_bitcoin_qr.png and /dev/null differ diff --git a/manuscript/recipies/cryptominer/nvidia-gpu.md b/manuscript/recipies/cryptominer/nvidia-gpu.md index efb32fd..6d7504c 100644 --- a/manuscript/recipies/cryptominer/nvidia-gpu.md +++ b/manuscript/recipies/cryptominer/nvidia-gpu.md @@ -49,10 +49,10 @@ reboot !!! warning Like overclocking itself, this process is still a work in progress. YMMV. -Of course, you want to squeeze the optimal performance out of your GPU. This is where the X11 environment is required - to adjust GPU clock/memory settings, you need to use the ```nvidia-settings``` command, which (_stupidly_) requires an X11 display, even if you're just using the command line. +Of course, you want to squeeze the optimal performance out of your GPU. This is where the X11 environment is required - to adjust GPU clock/memory settings, you need to use the ```nvidia-settings``` command, which (_stupidly_) **requires** an X11 display, even if you're just using the command line. -This command gives you a "fake" screen so that X11 will run, even on a headless machine managed by SSH only: +The following command: configures X11 for a "fake" screen so that X11 will run, even on a headless machine managed by SSH only, and ensures that the PCI bus ID of every NVidia device is added to the xorg.conf file (to avoid errors about ) ``` -nvidia-xconfig -a --allow-empty-initial-configuration --cool-bits=28 --use-display-device="DFP-0" --connected-monitor="DFP-0" +nvidia-xconfig -a --allow-empty-initial-configuration --cool-bits=28 --use-display-device="DFP-0" --connected-monitor="DFP-0" --enable-all-gpus --separate-x-screens ```