Getting started on Doca/GpuIO

Would like to experiment with packet processing capabilities and thruput testing with Doca/GPU. Ideally I would like the DPU to pre-process packets before I get Cuda kernels to work on them.
Whats the best way to go about experimenting/doing this?
I haven’t been able to see any “cloud” resources that I could rent with Doca/Bluefield nice to play with. Any pointers?
If I have to buy a server with Bluefield Nice and Gpu… can someone point me to a system I could get?