Skip to content

Commit ed7532d

Browse files
committed
Add documention on how to use
1 parent 2f810ed commit ed7532d

File tree

1 file changed

+14
-0
lines changed

1 file changed

+14
-0
lines changed

docs/docs/configuration/hardware_acceleration.md

+14
Original file line numberDiff line numberDiff line change
@@ -63,6 +63,20 @@ A more complete list of cards and ther compatible drivers is available in the [d
6363

6464
If your distribution does not offer NVIDIA driver packages, you can [download them here](https://www.nvidia.com/en-us/drivers/unix/).
6565

66+
### KHADAS VIM3 SBC
67+
68+
This SBC is equipped with the Verisilicon TPU. Khadas Website (https://www.khadas.com/vim3)
69+
config.yml
70+
```
71+
vim3:
72+
type: vim3
73+
```
74+
75+
NOTE! - most distributions do not include a working galcore.ko kernel driver that matches the NPU module. [Download galcore.ko](https://drive.google.com/file/d/1pxCDMp686kCuYvdABJsWFBVvXnMBXk0z/view?usp=sharing) and place in the /lib/modules/4.9.241/kernel/drivers/amlogic/ directory [Download link TBC]
76+
Debian11 build link
77+
78+
This needs to run in Docker --priviledged mode
79+
6680
#### Docker Configuration
6781
6882
Additional configuration is needed for the Docker container to be able to access the NVIDIA GPU. The supported method for this is to install the [NVIDIA Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html#docker) and specify the GPU to Docker. How you do this depends on how Docker is being run:

0 commit comments

Comments
 (0)