$AZTEC
Fading QuicklySnapshot Window: 2026-07-24 00:10 UTC ยท โ Back to Crypto Overview
Social Momentum Summary
Total Engagement - Comments: 0, Retweets: 1, Likes: 6, Impressions: 263
Verbatim Community Citations & Social Evidence 2 source posts analyzed
How it feels to be Stage 2
![]()
Onchain @perps at scale take more than low latency. They need predictable execution, deterministic ordering, and reliability under load, without compromising compatibility. With Parasol, we're extending the Agave client to build the highest-performance CLOB on Solana. Our CTO [Spoken audio]: We try to keep the compatibility with Solana as much as possible. And the main thing which we want to resolve in Paraso is the property of centralized solutions like deeper liquid for example or some L2 solution for Ethereum and so on. It's an endpoint to where you can go and the transaction confirmation you will receive faster than it's possible in the main chain and also you can predict how many time is needed for transactional execution. It's more critical. It's critical for traders when they make the decision to create some transaction and send it to the blockchain to which part of the block it will be included and how fast it will happen because the situation in the blockchain change very fast, and the speed of this length of transaction can be very critical in different scenarios. So, we developed this solution, which provides additional functionality for Solana, where you can build, centralize it and point of accept and transaction, allow you to be guaranteed that it doesn't depend on the distance to the block producer. And this is our purpose. And what we change directly in GavaNode, we change things which can be implemented in Solana mainnet too. For example, it is sub-accounts, it's technology, which allows to create on-chain routing of transaction flow. We don't break Solana possibilities to execute transaction parallel. We just introduce the functionality, which allows to split big count in small chunks. each chunk has its own address, and this address depends from the base account. You can't load these subaccounts directly in transaction, you can do it only from the SMM program. It means that you don't have conflict or execution on chain, and this allows to decrease number of accounts inside of transaction. It decreases amount of data which should be written to the account db at the block finalization time and other stuff is just optimization for how many data you should read for processing and how many data you need to write to the state. Also, we changed the mechanics for finalization. Why? Because we need them to create a state projection. It means that the transaction which we send to Parasol and to the Solana can use the same shared assets between them without anything to do on the user side. Everything happens on the parasol side. Next functionality is sorting of transactions, which can be required by the application because it is removed. For example, math mechanics, which are available in networks with priority fees. And also other stuff is just a simple modification, which can be required by different applications. But it's not so critical, like increase transaction size if application needed increased limits for transaction execution, but all these stuff can be required by different applications in different variants, but it's not something critical and difficult for modification in any fork of the line. So main components sub-accounts. Next one is rules for finalization. Next one is ordering of transaction. All other stuff is not so important like this reward.