Moonspark 🅿️

ᴢɪʀᴇɴ 2.0 ɪs ʟɪᴠᴇ, ᴀɴᴅ ɪᴛ sɪᴛs ᴀᴛ ᴛʜᴇ ʜᴇᴀʀᴛ ᴏꜰ ɢᴏᴀᴛ ɴᴇᴛᴡᴏʀᴋ's sᴛᴀᴄᴋ. @GOATNetwork says every state transition of its zkEVM is proved by Ziren, and that its BitVM3 bridge is designed to settle on those proofs. So when ZKM, the team behind Ziren, replaces its proving core, it matters to GOAT. Here is what changed in short. ᴡʜᴀᴛ ᴄʜᴀɴɢᴇᴅ A zkVM runs a program and produces a proof that it ran correctly. Ziren 2.0 swaps the engine that produces those proofs for a new proof system called jagged-over-WHIR, rebuilt across 1,690 commits and documented in a design paper on the IACR ePrint Archive. Instead of one separate check per table, one combined check covers every chip's rules, and the data is stacked at its real size with no padding. Commitments are hash-based throughout. Programs that built for earlier versions build unchanged, in Rust, C and Go. ᴛʜᴇ ɴᴜᴍʙᴇʀs - Speed and cost: on Ethproofs, the Ethereum Foundation's open proving dashboard, Ziren proves the average Ethereum mainnet block in 8.6 seconds on a four-GPU cluster, at an average of $0.0061 per proof (as of Nov 3) - Size: the final proof is about 284 KiB, however long the execution ran - Checking: native verification takes about 25 milliseconds - Scaling: a 358-million-cycle block took 51.3 seconds on one GPU and 16.7 on four - Range: fifteen consecutive mainnet blocks proved end to end with two different execution clients, Reth and Geth The path so far, per ZKM: production 1.0 in May 2025, an independent evaluation by ProofLab in September 2025, Ziren becoming the first zkVM to prove Geth without modifying the Go compiler in April 2026, the design paper in October 2026, and now 2.0. ᴛʜᴇ sᴇᴄᴜʀɪᴛʏ ᴀᴄᴄᴏᴜɴᴛ Determinism is machine checked in Lean: 57 chips, 104 theorems, all closed. The paper states more than 100 bits of interactive soundness at every proving stage and 93.5 bits of composite security for a whole block, and it lists the open items in one place. Soundness rests on the verifier and what it reads. A bug in the executor or GPU kernels yields a proof that fails to verify, not one that verifies falsely. ᴡʜᴀᴛ ɪᴛ ᴍᴇᴀɴs ꜰᴏʀ ɢᴏᴀᴛ ɴᴇᴛᴡᴏʀᴋ - Cheaper, faster proofs matter to a network that proves every state transition - A small, fixed-size proof is quick to check, which suits a proof-based bridge - Security you can inspect: public benchmarks, Lean checks and published accounting - A route toward Bitcoin: Ziren is developing a hash-based verifier for BitVM-style protocols, where a Boolean circuit checks the proof on Bitcoin only if a dispute arises - Less friction: existing programs carry over That hash based route also matters for GOAT's post-quantum plan. The pairing based wrapper is one of the surfaces GOAT's quantum-exposure inventory lists as exposed, and a hash based verifier is the intended replacement. ʜᴏɴᴇsᴛ ᴄᴀᴠᴇᴀᴛs These benchmarks are for Ethereum blocks, not GOAT's own workload, so GOATNetwork's numbers will differ. BitVM3 is still in development. And the figures come from ZKM's own paper and documentation, though the Ethproofs listing is public. A team that has spent years putting security ahead of speed now has public speed numbers too. Details: docs.zkm.io