Equilibrium Group

This first of its kind report by Hannes Huitula from Equilibrium Labs is mapping the Web3 confidential computing landscape, and featuring @sodabubblelabs' #gcEVM! The thesis is: 'All roads lead to MPC,' as various cryptographic techniques (ZK, FHE, GC, SS) ultimately converge into MPC protocols. https://lnkd.in/dzG4Bp9D

Do All Roads Lead To MPC? 🚧 Our latest research piece argues that if the desirable end-state is programmable privacy infrastructure that can handle shared private state without any single point of failure, then the trust assumptions of all approaches boil down to those of MPC 🔍 We also explore the maturity of MPC and its trust assumptions, highlight alternative approaches, compare tradeoffs, and provide an industry overview. https://lnkd.in/d-ytEVDc

🚄 Equilibrium Labs Infra Bulletin #30 covers Obol's release of Charon V1, Geographical decentralization of Ethereum, Reth 1.0, and more… Let's dive in! 🤿 Highlight of the week 🔍 Obol released a production-ready version of their DVT middleware, Charon. Distributed validator technology (DVT) enables multiple nodes to collaborate on running an Ethereum validator ("multi-sig for validation"). Charon is the middleware client that sits between the validating client and its connected beacon node, enabling existing Ethereum validator clients to operate together as part of a DV cluster. Research, Articles & Other Things of Interest 💡 - Research on the geographical decentralization of Ethereum Nodes - Research on MegaETH and their design choices - Paradigm releases Reth 1.0 - Vitalik's recent post around pre-confirmations - ZK FM podcast episode on programmable privacy - Demo of Starkware's new Stwo prover (1,000x improvement) Personal Recommendations From Our Team 🤌 📚 Seeing Like A State - James C. Scott 🎧 Bleeding Me Dry - Alicia Creti 💡 Faces Made Of Living Skin Make Robots Smile Link to newsletter in the comments - thanks for reading! 🙏

🚄 Equilibrium Labs Infra Bulletin #29 covers Pre-confirmations, RiscZero's zkVM 1.0, Privacy in Blockchains, and more… Let's dive in! 🤿 Highlight of the week 🔍 Preconfirmations are promises (but not guarantees) of future execution or inclusion. They aim to increase the user experience in cases where final execution might take longer. There are many ways to implement them though in blockchains and each one inhibits tradeoffs. The article by Longhash offers a comprehensive overview of the current space. Research, Articles & Other Things of Interest 💡 - Privacy in blockchains - what, why, how - Can blockchains and ZK help humanity survive? - RISC Zero launched their production-ready zkVM and published performance benchmarks. -  Tradeoffs with zkVM design Personal Recommendations From Our Team 🤌 📚 Chaos: Charles Manson, the CIA, and the Secret History of the Sixties - Tom O’Neill 🎧 Harpists have 7 pedals - one for each note 💡 Joining an "AI" startup Link to the newsletter in the comments - thanks for reading! 🙏

ANNOUNCING: EQUILIBRIUM VENTURES (EQV) - An Engineering-first Venture Firm 🎉 I'm stoked to share this: We've launched our first venture fund, Equilibrium Ventures (EQV), supported by 70+ sr engineers who have helped build core infrastructure for major blockchains, including Starknet, Ethereum, Solana, Aptos, Polygon, Celestia, zkSync, Aleo, Polkadot, Ripple, and Algorand. Huge congrats to m honkasalo, Henrik Sundvik, and Chris(Topher) A. for making this happen. tl;dr this fund 🔬 conducts deep technical DD in blockchain infra and cryptography, and 🏗 actually helps founders build. Curious to know more? Check out the comments for our recent features 👇

🚄 Equilibrium Labs Infra Bulletin #28 covers Deep Reals, MEV on L2s, Blended Execution and more… Let's dive in! 🤿 Highlight of the week 🔍 A deep real is an inversion of a deep fake. This "fake until proven real" mindset is related to the attestation problem, which is only getting more acute with generative AI and marginal cost of producing content reaching 0. We're all for more decentralized means of providing "trust points", but these can be implemented in different ways. Blockchains may be part of the solution, but it's not the solution by itself. Research, Articles & Other Things of Interest 💡 - MEV on L2s (a paper and a separate podcast on the same topic) - Polygon using Succinct's generic zkVM SP1 for the AggLayer - Starkware announced their expansion to Bitcoin (conditional on OP_CAT passing) - Fluent is aiming for blended execution, where apps on different execution environments can call each other seamlessly (WASM, EVM and SVM to start with) News From Our Partners 🔥 Aligned Hauz is a hacker house during ETH CC that is open for both residents and visitors. For us, "unaligned" is a play on the tribalism in crypto - to be unaligned is to be open to new ideas. Organized by EQ's very own Joakim Isoaho and a few others in the industry. Personal Recommendations From Our Team 🤌 📚 The Book of the New Sun - Gene Wolf 🎧 Minoru Muraoka - Bamboo 💡 When Privacy Expires Link to newsletter in the comments - thanks for reading! 🙏

🚄 Equilibrium Labs Infra Bulletin #25 covers Enhancing Client Side Proving, GPU-EVM, State of Light Clients and more… Let's dive in! 🤿 Highlight of the week 🔍 Penumbra's recent upgrades aim to speed up client-side proving, where rather than outsourcing to a specialised third-party, users generate the ZKPs locally (on their own device) which maximises privacy. While some efforts are being made around outsourcing private proofs (with the help of MPC), client-side proving is expected to remain an integral part of privacy protocols. Hence, any improvements in speed and performance of these proofs significantly improve the experience for the end-user. Research, Articles & Other Things of Interest 💡 GatlingX announced EVM-GPU (a highly parallellised EVM implementation), our "2024 state of light clients" covers existing light client implementations, a report on FHE highlights 5 big outstanding problems to solve, and more... Personal Recommendations From Our Team 🤌 📚 How bad policy favors memes over matter - Chris Dixon's recent op-ed 🎧 GoGo Penguin - From the North 💡 The passing of the US surveillance bill is bad news for privacy advocates Thanks for reading - link to full bulletin in the comments ⚡️

State of light clients in 2024💡 With a growing number of light client implementations, it can be difficult to stay up to date with the latest developments. We spent some time with Joakim Isoaho looking into the space and wanted to share our findings below (link to full table in comments)👇 We aim to keep updating the table over time, so please comment below or DM if we missed any projects! Light clients enable verifying some aspects of the underlying blockchain without having to download all the underlying data and re-execute, hence requiring much less resources than full nodes. This is useful for various use cases, such as enabling easier/quicker syncing, bridges utilising light clients, and verifying that data was published correctly with data availability sampling (DAS). However, not all light clients are the same and ideally, you want a "full" LC that verifies consensus, DA and correct execution. The holy grail is to get as close to full node security as possible, without the high hardware requirements. Most LCs today only verify consensus though, and some only a subset of it. Some observations from the table: 1️⃣ Most light clients today only verify consensus (and some only verify signatures from a subset of validators, such as light clients using Ethereum's sync committee). Full LCs with execution verification and DAS are likely still too expensive and to some degree, would require changes to the underlying blockchains. 2️⃣ Ethereum LCs rely on the Altair Sync Committee - a small subset of validators which is rotated every 27h. This has problems both due to lower cost to corrupt and misaligned incentives with no slashing of the sync committee. There is a proposal (EIP 7657) that's focused on improving the sync committee by adding slashing to better align incentives. Other teams are working on implementations that include the whole validator set. 3️⃣ Most light clients are developed by external teams (both due to resource allocation and more aligned incentives). Developers of light clients can be divided into three groups: 1) Core team 2) External team with own incentives (for example a bridging protocol) 3) External team doing public goods research (e.g. a16z with Helios) Thanks for reading ⚡️