Requirements clarification: Anbox Cloud worker nodes on Rockchip RK3588 (8 cores, 16GB, Mali-G610 MP4)

The requirements page lists arm64 worker nodes at a minimum of 8 cores and 16GB RAM, and notes that virtualized Android images additionally require host hardware virtualization with /dev/kvm available. I would like to confirm whether a specific arm64 platform qualifies before we invest in a trial.

Per node:
- Rockchip RK3588: 4x Cortex-A76 plus 4x Cortex-A55, 8 cores total
- 16GB RAM
- Mali-G610 MP4 GPU
- NVMe storage
- 1GbE
- /dev/kvm present and functional

Questions:
1. Is a node sitting exactly at the documented 8-core and 16GB minimum a usable production worker, or is that floor intended for evaluation only? What container density would you expect per node for containerized Android at that spec?
2. Is the Mali-G610 MP4 a supported rendering resource? I want to confirm whether a Panfrost path or a Mali blob path is expected, or whether we should plan on software rendering for headless CI workloads.
3. For virtualized Android, does RK3588 satisfy the hardware virtualization requirement in a way Anbox Cloud can consume? It is ARMv8.2 without nested virtualization, and I want to know if that is a blocker.
4. Is there a known constraint on running the Charmed distribution across a large number of small worker nodes rather than a smaller number of large ones? We are looking at roughly 1,280 nodes at the spec above.
5. What does licensing look like for a deployment of that node count, and is there a route for an infrastructure owner rather than an end customer?

Start with the requirements page referenced in the issue; no file path, test, or entry point is provided. Determine whether maintainers want documented answers about RK3588 support, rendering, virtualization, node scaling, and licensing, then define completion as an authoritative update or response covering the applicable questions.

Written by the indexing model from the issue text.