krak/

Thoughts on the ATLAS launch by @LayerZero_Core The team has been working relentlessly since the Zero announcement back in February. ATLAS is really the first real look at how the team is methodically attacking what I think are the biggest problems L1s have today. Much more will come, but on performance: these are internal benchmarks, but Zero is already operating at better performance than crypto CEXes. In a non-colocated environment, ATLAS reports <1ms median latency, 1.4ms p95, and 2.6ms p99, measured wire-to-wire from receiving an order through full execution and putting the response back on the wire. For comparison, optimized/colocated measurements of Binance are generally in the low single-digit milliseconds, with third-party benchmarks around ~2ms median and ~8ms p99. More will come on this as they release more info. What I'm genuinely excited about is the fee structure and incentive design. It's becoming abundantly clear you need a different set of participants to make trading venues successful at scale. We've discovered this slowly in crypto, but it's becoming very clear with Hyperliquid and HIP-3 markets. Zero is being deliberate about how it makes trading activity flourish, with rebates to venues, exchanges, and brokers, but also to market creators (which sometimes collapse into the same entity). If you want to house serious financial activity, you need other people making money alongside you. Otherwise you end up in direct conflict with your own ecosystem, or everyone just forks you. In this model there's one unified fee the user pays, where app-level fees and gas collapse into a single number. That's better for the user: more transparent, more predictable. For exchanges, venues, and market creators, there's a clear revenue stream and value accrual. Same for ZRO holders. My biggest takeaways, basically four things I think L1s have not been addressing: 1) MEV misalignment and broken L1 economics: L1 base fees are trending toward zero, and priority fees/tips only get generated at scale when there's real contentious state to bid on. The clearest example is Solana: it's at an all-time high in transaction count, yet making yearly lows in revenue. Apps are internalizing most of that MEV themselves, because they also want revenue and want to protect their own users. This creates real misalignment, extractors want retail to get wrecked so the L1's top line looks good (REVoooors). That is not sustainable. 2) ATLAS is fully neutral infrastructure. The Zero team has said they will not launch a competing exchange or frontend, and they've been working hard to onboard others instead. The problem with HIP-3 markets is there's a real misalignment with the main Hyperliquid exchange: fees stack on top of the exchange's own fees, and Hyperliquid has an incentive to list things itself before HIP-3 markets get the chance, to capture more economics. Your infra provider ends up competing with you for flow. 3) Kill the brainless 100% automatic buyback and burn. The rebate mechanism and tri-party fee distribution give every party skin in the game to grow volume, while still giving the Zero team room to tune the model as needed. It also just doesn't make sense to charge CME the same fee you'd charge a small broker. 4) ZRO holders accrue value directly from the app layer, since everyone takes a cut of one unified fee. The team is maximally aligned with the user, nobody is pushing MEV-driven narratives just to make the charts look good while users get burned underneath. This is a first step, and mainnet is always messier than the deck promises. Still a long way to go.