The data growth of Kaspa

These are some rough estimations of the storage needed to run full nodes / pruning / archival.
Please note that the numbers are still subject to change due to main net not being out yet.

Note that our Difficulty Adjustment Algorithm promise us a total block rate of 1 block per second. (not 1 blue block or 1 red block. but overall 1 block per second).

Full pruned nodes need to keep:

  1. All past headers.
  2. The full UTXO Set.
  3. All blocks blocks data until the pruning depth which is around 3 days (244,838 in blueScore difference)

Archive Nodes:

  1. Everything that Full pruned nodes keep.
  2. All past data, including before the pruning depth.

For pruned nodes this mean:

  1. Headers - Minimum of ~4.8GB a year, Maximum of ~14GB a year (depending on the width of the DAG).
  2. The UTXO set size depends directly on the amount of users and the way they use the system, for reference Bitcoin’s UTXO set is currently ~4.3GB We might have a slightly bigger UTXO set for the same usage and users, because we have more coinbase transactions, but it’s really hard to predicate how often will miners consolidate their outputs (if they don’t it will grow by ~1.7GB per year).
  3. If we assume negligible amount red blocks (which is correct assumption for 51% honest and working system), then this accumulates to maximum of ~25GB.

For archive nodes:

  1. Same as above.
  2. For block size limit of 100KB this will get to around ~3TB a year maximum (note that this assume very high tps).

Can you please demonstrate how you got to these numbers? Specifically, the block headers