Its an exciting day! After months of work I’m glad we’re finally able to share details of the work we’ve been doing and is going live in the @zksync Atlas upgrade. I figured I’d take a minute to share some more info on how the aspects of this upgrade come together. Importantly, we’re building for a future in which the world is onchain. Crypto systems will be the backbone for the vast majority of the world’s financial systems, which is is precisely why incorruptible financial systems are required. To me, this onchain economy has to be reflected in the architecture. I expect enterprises, banks, manufacturing facilities, hospitals etc to interact with the onchain economy via infra they’re operating, just as they manage production software stacks for their other operations today. Teams will customize their systems to meet their privacy, regulatory, or user experience needs. A simplified version looks like this: Where we have many ZK chains, some private (red) and others public, configured to support their specific use case. Every second these chains will be generating a proof for their system, which gets verified on the ZK coordination layer. Within one second of the user sending a tx, it has this ZK finality which is automatically verifiable by the entire network. Fast ZK proofs are a breakthrough technology. With this, you solve the problem of chains not being able to coordinate effectively. If a user wanted to send 100 USDC from their public account on one ZK chain to a friends private account on another ZK chains - this is immediately possible. In this example, ZK chain 5 doesn’t have to “trust” that ZK chain 2 is behaving honestly with the transfer. The ZK proof (represented by our little ZK safe) proves that the message received by chain 5 can be acted on without concern. For this vision of an incorruptible global network, connectivity is clearly a critical aspect. The second critical piece is the system you operate as part of this network. Engineering teams want to be deploying software that is performant, reliable, secure etc. This is system that provides the foundation for your UX. As part of the Atlas upgrade we’re deploying a brand new version of the sequencer, designed from the ground up for low latency tx inclusion (so the system feels fast to your users) and for throughput that can bring enterprise level activity onchain. This is made possible by an extensive upgrade to the “operating system” of the ZK Stack chains, including changes to how we think about what should be proved. With Era, we essentially had the VM implemented twice: once in Rust code (that was recognizably a VM) and once in the Boojum cryptographic circuits. With Atlas, we’re taking a more general approach and are now proving RISC-V. This, along with changes to how we organize the execution of transactions, is how we’re able to support the EVM natively, but with a system that can flexibly support other VMs without modifying the proof system. If it complies to RISC-V - you can prove it! These changes are like a massive engine upgrade to the ZK Stack, and together provide the tooling to build and operate systems that are more performant and more connected than has ever been possible before. More to come. I want to finish with a huge thank you to everyone that has worked on this upgrade! A huge amount of effort has gone in to bring this to life. ♥️