Ep. 743 Why Monad Is Taking Ethereum’s Biggest Bottleneck Head-On with Keone Hon

Keone Hon co-founded Monad with James Hunsaker, whom he met at Jump Trading in 2014 and describes as the best engineer he ever worked with. Hon spent roughly eight years at Jump starting in 2013, where pipelining was a core design principle in building high-performance trading systems that competed on major futures exchanges including CME, EUREX, ICE, and the Osaka Securities Exchange. He moved into crypto at Jump in 2020 and co-founded Monad because he believed nothing available in the market matched what they could build themselves. Hon acknowledged that HFT experience does not straightforwardly translate to blockchain development, since centralized infrastructure co-located near exchanges differs fundamentally from a decentralized validator network with unknown hardware and locations, but he views blockchain performance as fundamentally a technology problem with meaningful overlap in insights from high-frequency trading.

Monad is an EVM-compatible layer one blockchain, not a layer two. Hon chose EVM compatibility because it is the dominant standard with the most builders, tooling, network effects, and applied cryptography research. Ethereum currently has 12-second block times, roughly 20 transactions per second, and finality taking approximately 12 minutes. Monad produces blocks every 300 milliseconds, supports 10,000 transactions per second, and achieves finality in 600 milliseconds, making its blocks over 36 times more frequent than Ethereum's and its finality roughly 1,440 times faster.

Monad achieves its performance through two distinct parallelization techniques. Pipelining runs consensus and execution simultaneously rather than sequentially, analogous to interleaving loads of laundry rather than completing each one before starting the next. Monad also executes transactions in parallel within each block. Consensus refers to getting globally distributed nodes to agree on an official transaction ordering, while execution refers to running the virtual machine that produces the resulting state changes. Monad recently published a paper describing a new consensus mechanism called Cadence that enables multiple concurrent proposers, achieving 100 millisecond block times with 200 millisecond finality, which Hon says approaches the physical limit imposed by global roundtrip transit times of several hundred milliseconds.

Pre-trade privacy is a major unsolved problem on existing layer one blockchains, where mempool transactions are visible and subject to front-running and sandwich attacks that result in MEV extraction. Hon stated that no blockchain has solved the MEV and mempool privacy problem to date. Monad blocks are shielded until they are committed, which addresses part of this problem. Monad also published a second research paper describing BTX, a batch threshold encryption protocol that enables encrypted mempools so submitted transactions cannot be front-run before inclusion.

Liquidity quality on Ethereum is poor in practice because high transaction costs prevent market makers from updating quotes frequently, forcing them to post wide passive orders far from fair value, meaning nominal liquidity figures are misleading and slippage in practice is massive. Faster finality on Monad enables trading systems, exchanges, money markets, and liquidation systems to deliver a meaningfully better user experience. Purple, a fully on-chain perpetuals DEX on Monad, traded 300 million dollars of volume in the past seven days, with every order add, cancel, and trade recorded as an on-chain transaction made possible by cheap fees and fast finality. Kuru, a fully on-chain spot order book DEX on Monad, trades approximately 20 to 25 million dollars of volume per day. Sub-one-second finality also meets the threshold payment networks require to process credit and debit card transactions on-chain, since on slower blockchains card issuers must tie up large amounts of capital to front settlement risk.

Running a Monad validator requires hardware nearly identical to an Ethereum validator, with the only difference being a slightly larger SSD costing around 200 dollars. Hon stated a Monad validator can be run on a Costco MacBook. Monad is proof of stake, with node operators staking to their own node and earning rewards proportional to stake, and delegators able to stake to a node operator who earns a portion of those earnings. Monad currently has 200 nodes in production, fully globally distributed, with node distribution publicly visible at gmonads.com. Hon noted that most blockchain foundations do not advertise their node distribution because it is centralized. Monad was the first project on Coinbase's new token sale platform, giving retail users access to the token ahead of public mainnet launch, with the foundation prioritizing broad distribution to create many stakeholders aligned with network success, citing Bitcoin and Ethereum as models for that approach.