eigenlayer

Wrote a small primer on Espresso's shared sequencer (SS) and how Eigen interacts with it.  www.eigenhub.org/avs-hub/blog-post-title-one-5876y-cssaa 
Check it out! I hope you find it helpful as we all think about node ops/restaking for SSs in the future. 2 other SSs Aztec Fertnet and Astria articles to come soon. 🧵

Two eigen enthusiasts looking to share all things eigen check out https://www.eigenhub.org/avs-hub
or stake with us here 
 https://www.eigenhub.org/

-Sequencers are a necessary piece architecture needed for L2 operation. Most sequencers are centralized, bad for security and decentralization. Afaik Arbitrum and Optimism still just have one completely centralized sequencer, but are working to change that with Eigen's, and your help.

-With Espresso’s Sequencer and Eigen joining forces roll-up sequencers can have it all, high throughput and low latency with Eth decentralization and security through Proposer Builder Separation. DA and consensus become the focus of Espresso and Eigen nodes handle execution. A winning combo!

You can see in Eigen Hubs Quick Guide (a work in progress as we learn about Eigen!) that solo node ops with low end hardware might should focus on the shared sequencer AVSs coming. The hardware load shouldn't be heavy since you run execution only, think litenode. Espresso handles the DA/consensus easing your burden

Referencing above, we are not sure what APY restakers can expect yet. No AVSs on the market. However, this is crypto, so you know we got to speculate! With the main trust levers being economic and fairly decentralized node set, which needs to be incentivized with re-staker TVL, I would expect a mid range APY for SS.

Competition will also be a driver several SS AVS in the pipeline. I'd guess 10% - 20%. Again, dubious speculation!  Please let me know if you have  corrections/thoughts. It is an exciting time in the Eigen Space and Eigen Hub wants learn and grow right there with you!

-Sequencers should have high throughput and low latency, but also be decentralized, have low trust assumptions, and be secure. The first two are hard to implement without compromising the others.