Resident vs Non-Resident
conceptBoth are hardware-backed SSH keys with the same cryptographic strength. The difference is what the hardware remembers: a resident key is generated and stored entirely inside the device; a non-resident key is derived on demand from a blob file on disk. Resident vs non-resident is about state survivability, not strength.
The critical summary:
| Feature | Non-Resident | Resident |
|---|---|---|
| Secrets on disk | ❌ | ❌ |
| Blob required | ✔ | ❌ |
| Stub regenerable | N/A* | ✔ |
| Uses slots | ❌ | ✔ |
| Unlimited | ✔ | ❌ |
| Disk loss safe | ❌ | ✔ |
| Best use | GitHub | Servers |
*Non-resident keys use blobs (required, not regenerable), not stubs. Only resident keys have stubs, which can be regenerated on any machine.
Neither puts a secret on disk. The question is which non-secret file sits there—a blob you must preserve, or a stub you can throw away—and whether the key consumes a scarce hardware slot.