Discussion Details
Starstream: next gen VM
govtooladmin
Description
Starstream is a new virtual machine (VM) that iterates on all the lessons learned from Plutus to build a new VM for Cardano. Instead of the future of Cardano relying solely on iterative improvements to the Plutus design, Starstream is a radically different approach at designing a utxo-based VM that maintaing all key properties such as determinism.
Notably, it differs from Plutus in a few key ways:
- Starstream is designed using coroutines as the core primitive. Everything including transactions, tokens, utxos and more are modeled as coroutines. Having a single core primitive to model all constructs leads to a simple VM design
- Starstream is designed with ZK cryptography in mind. This means that not only is it simple and efficient to generate ZK proofs of Starstream, the proofs properly compose with all concepts including full transactions. This means that, by leveraging ZK, Starstream supports batching arbitrarily many things together into a single batch proof unlocking a lot more scalability
- Starstream supports computation over private data. This means, for example, you can prove you do not live in the US without having to reveal your identity, or prove you own an NFT in a collection without having to reveal your address.
- Starstream supports onchain ABI for contracts improving composability with both onchain and offchain constructs (unlike Plutus which is untyped onchain)
Starstream is also designed with upgradability in mind. Unlike other ZK-based systems that need to be fully rebuilt every time a new research breakthrough comes out, Starstream is instead built as a way to add a utxo layer on top of an existing general zk proof system. That means that when new better proof systems are developed, Starstream can support them with minimal effort and without the change causing cascading disruption to other parts of the stack. Technically speaking, Starstream does this by being built solely as a layer of additional memory consistency checks that can be placed on top of another zkVM construction
Problem Statement
Build the initial implementation of Starstream - a new VM for Cardano that improves on the shortcomings of Plutus while being built to support ZK cryptography directly on the L1
Proposal Benefit
Starstream supports use-cases that require private computation (necessary for many RWA use-cases) enabling a new category of apps to exist on Cardano. It does this while also allowing for near-zero cost of batching txs which massively improves scalability even for DeFi use-cases
Key Proposal Deliverables
milestone 1: a developer playground for people to try Starstream from their browser on a mock ledger
milestone 2: a governance info action to see if people agree on the direction and value of Starstream
milestone 3: integration of Starstream into at least one Cardano node implementation, and the launch of a testnet using it
milestone 4: a governance info action so see if people want Starstream added to the Cardano L1
milestone 5: integration into cardano-node
Cost Breakdown
While flexible on the contract type, this project does involve a certain amount of experimental research and cooperation between multiple parties. The cost breakdown amounts to 3〜5 FTEs for a year which should be sufficient for the developer playground and tentatively the testnet (but exact cose will depend on the node team we work with), and does not include the cost of cardano-node integration
Resourcing & Duration
milesone 1: 3 fulltime developers. PoC at the end of summer 2025 and playground in fall 2025
milestone 2: one month after milestone 1 (duration of info action)
milestone 3: 3〜5 fulltime devs (3 for Starstream, 2 for the node team). PoC in Q1 2026 and testnet launch in Q2
milestone 4: one month after milestone 2 (duration of info action)
milestone 5: TBD (depends on cardano-node team at the time)
Experience
Our team members and leadership have contributed to core tooling for Cardano for years, including having built the first EVM layer for Cardano (Milkomeda)
We also have multiple years experience in ZK cryptography having contributed to both Midnight and Mina
Maintenance & Support
Similar to development of Plutus, it will be a VM integrated into the chain and updated through protocol upgrades, CIPs and other existing core upgrade processes
Supplementary Endorsement
You can find the announcement of Starstream in this Twitter thread: https://x.com/sebastiengllmt/status/1898226507874697499
It includes a link to the Starstream Working Group (https://discord.com/invite/9eZaheySZE) whose members include representatives of all founding entities as well as many of the large infra and dApp companies working in the Cardano ecosystem
Roadmap Alignment
Does your proposal align with any of the Intersect Committees?
Technical Steering Committee
Does this proposal align to the Product Roadmap and Roadmap Goals?
Programmable Assets
Administration and Auditing
Would you like Intersect to be your named Administrator, including acting as the auditor, as per the Cardano Constitution?
No
Ownership Information
Submitted On Behalf Of
CompanySocial Handles
https://x.com/sebastiengllmtKey Dependencies
eventual support by one of the Cardano nodes (cardano-node, Amaru, Dingo, etc.)
Cast Your Vote
Comments (1)
It sounds good, but I'm trying to identify the downsides of this proposal, especially the technical ones. I'll spend a bit more time reading through X's tree later. Please comment if you see any concerns with this proposal.
Are You Ready to Participate?
Building Together to Drive Cardano Forward.