Is there real tenant demand for arm64 CPU capacity, and what does a provider at that spec actually earn?

[Josh-Difratech]

Considering onboarding a large arm64 bare-metal fleet as a provider, but I'm not sure if I can.

What we would bring:

• Roughly 1,280 arm64 nodes, 8 cores and 16GB RAM and NVMe each
• US Tier III datacenter, single site initially
• Native arm64, not emulated

Questions:

1. Does the marketplace see meaningful bids for arm64 leases today, or is deployment demand effectively x86 and GPU only? Any actual arm64 lease counts would help a lot.
2. For CPU-only capacity at that spec, what is a realistic monthly revenue per node? I am comparing against a fixed colo cost, so even a wide range is useful.
3. What is the operational burden of running a provider this size? We do not have engineering staff. If this needs a full-time operator, it is a non-starter and I would rather know now than after onboarding.
4. Is there a minimum node spec below which nodes are effectively unbiddable?