Reproduced from this thread. “FALSE SEQUENTIALITY” credit to Danny Ryan.

The eth1 >> eth2 naming convention implies SEQUENTIALITY: aka things are strictly ordered. In reality, there will be some overlap during the transition — and that’s ok. #serenity (disclaimer: phases not set to a timescale, this is a possible future, not official)

Key Points: — ETH1x: existing user data, your favorite dapps will continue to exist unchanged until Phase 1.5. — Deposit contract state bootstraps Phase 0 Genesis from ETH1x

Phase 0: Once enough ETH is sent to the contract, the Beacon Chain will begin. Validators will only be responsible for running the BC itself. Client teams will use this time to harden their clients, improve performance, and start implementing future phases.

Rollups can leverage ETH1x for Data Availability ↑

Phase 1: Shards are added! But they can only host data, not execution, aka sending transactions like we are used to on ETH1x. This is perfect for Rollup constructions, which will happily anchor to the Beacon Chain tp use this space ↑ ↑ ↑. Offchain scaling FTW!

Phase 1.5: At this point, PoW and miners on ETH1x will be replaced by the existing Beacon chain PoS and Validators. Research for “Catalyst” is being spearheaded by Guillame Ballet (@gballet) and Mikhail Kalinin (@mkalinin2). See the latest summary from ETHOnline.

Phase 2: Existing Shards gain the ability to execute transactions natively ↕ ↕ ↕ . Research continues on the optimal integration of this feature.

See this post from @VitalikButerin suggesting possible use of existing paradigms: https://ethereum-magicians.org/t/a-rollup-centric-ethereum-roadmap/4698

.. and another framing from @lightclients: https://lightclient.io/blog/eth2-is-a-rollup/

In the future, we’ll be able to drop all of these confusing names and Phases and it will all just be called Ethereum. I hope the diagram is helpful in putting all of these terms in context with each other.