Acne Buterin in comments on Reddit called on the developers of clients to begin work on the first phase of Ethereum 2.0. He noted that the specifications required to run the first phases have already been explored and now only requires optimization.
I would say that now there are no pending studies for the first phase; now it is a question of optimization and development of specifications. I actively insisted to the team that is developing a client, began work on the first phase as soon as possible (so we can better understand how the first phase works in practice).
I believe that the specification of the first phase is quite simple ― at least conceptually, most of the complex issues involved in this phase have been resolved to zero. The main technical risk in phase I (from my point of view) is that sharding p2p network has never run this way, and therefore we run the risk of “unknown unknowns”. However, we use sharding p2p network that has limited bandwidth, so we mitigate these technical risks.
My main task is to study the possibilities of starting a light client inside eth2 eth1 (or bridge) simultaneously with the first phase, because it becomes useful for rollup solutions on eth1 that will provide us with the promised scalability to 100 000 transactions per second; over the exact details of how this is done, still have to work.
In early June Buterin said that technology zkRollup will become “the dominant paradigm for scaling at least a few years.”
A zero phase Ethereum 2.0 aimed at launching the Proof-of-Stake without any other economic activity. The first phase focused on the launch of the basic model of sharding. During this phase we will work in parallel to the 65 “mini-blockchains” ― Beacon chain of zero-phase and 64 new shard.
See also: Overview of Ethereum 2.0 from BitMEX Research