We have Docker images ready with XRT installed, which can be used as base images to run FPGA applications. Here is an example of creating the pod with XRT Docker image.
$ cat << EOF | kubectl create -f -
apiVersion: v1
kind: Pod
metadata:
name: xrt-sample
spec:
containers:
- name: xrt-sample
image: xilinx/xilinx_runtime_base:alveo-2022.2-ubuntu-18.04
imagePullPolicy: IfNotPresent
command: ["/bin/sh"]
args: ["-c", "while true; do echo hello; sleep 3600; done"]
resources:
limits:
amd.com/xilinx_u200_gen3x16_xdma_base_2-0: 1
EOF