A world computer worthy of the world’s money

You can also read this post on my blog.
The goal of Monad is to do for the financial system what the internet did for information. We have built a very strong technical foundation so far, but what is most exciting is what that foundation now enables. The next phase of Monad's development solves problems that are preventing existing blockchain systems from meeting user and business expectation. If we solve them, we will deliver a world computer worthy of the world's money.
Our 'why'
The internet made information available to anyone regardless of their country or economic status. We want that same freedom to extend to money, where everyone gets access to the same opportunities and markets.
That requires a financial system that people anywhere can access, and that developers anywhere can build on. Assets, protocols and fintechs should be able to work together in one shared system. People should retain control over their money, and new businesses should be able to compete without asking permission from the companies they challenge.
As an industry, we have made progress toward that goal. However, there are significant shortcomings. MEV is treated as an inevitability, and block building is highly centralized on most blockchains. Key application logic frequently lives off-chain. And while traditional financial institutions are recognizing the value of blockchain technology, many are tempted toward private blockchain solutions as a means of addressing their requirements of confidentiality.
We have become too willing to accept these failures as permanent features of crypto. Users pay the cost every day. We should expect more from the infrastructure.
Our approach with Monad has been to build the architecture needed to address these problems. Moving more application logic on chain requires cheap computation and state access. Keeping transactions encrypted until finalization requires ordering them before execution. The first phase of Monad established those foundations. We built parallel execution, asynchronous execution, JIT compilation, MonadDB, MonadBFT, and RaptorCast. Together, these make it possible to do more work and reach agreement quickly across a distributed network.
With that foundation in place, the next phase is to deliver the protections and capabilities it makes possible. A user should be able to send an instruction without exposing it to someone who can trade ahead of them. An institution should be able to settle privately. A developer should be able to choose how an application orders transactions while connecting it to a common settlement layer.
Our ambition is a world computer capable of real world scale and execution quality. We are building it expeditiously.
What the first phase makes possible
A transaction requires several kinds of work. The network must distribute and agree upon ledger additions, execute instructions, and update the relevant state. Monad lets more of that work happen concurrently.
Parallel execution uses multiple threads and processors while preserving the agreed order of state changes. Asynchronous execution separates transaction ordering from execution, so consensus can advance while the execution engine processes earlier transactions. JIT compilation turns EVM bytecode into native machine code.
MonadDB supplies efficient access to state. With MIP-8, its page-aware storage tries commit to groups of consecutive storage slots, bringing the commitment structure closer to how disks read data. RaptorCast shares the work of distributing block data across validators. MonadBFT provides 300 ms block times and 600 ms finality in normal operation.
Getting to this stage has created opportunities for the next phase. Separating ordering from execution gives us a place to order encrypted transactions before opening them. Efficient computation and storage make it practical for applications to do more work on the user's behalf. A network that already shares data distribution work can take concurrency further in block construction.
The direction can be expressed through the properties we want users and developers to have:

From a visible order to a protected intent
Consider what should happen when someone wants to buy an asset.
They should be able to tell an application what they want to buy, how much they want, and the maximum price they will accept. The application should find a good route using the market state available when the trade executes. Sending that instruction should not give another participant an opportunity to exploit it.
Two problems stand in the way. First, almost all aggregators implement their routing logic off-chain - thus making routing decisions based on information that is already out of date when execution begins. Second, an observer who can read a pending order can react before its position is fixed, front-running the user's intent.
MEV is a powerful centralizing force. In Rated's 30-day view of Ethereum's builder market, the top two builders account for 72.9% of tracked blocks, and the top three account for 88.8%, as of October 3, 2026. Access to valuable exclusive order flow helps builders win blocks, which makes them more attractive to order-flow providers and harder for new participants to challenge. A distributed validator set alone does not prevent this concentration. We need to address the incentives that produce it.
Encrypted mempools address a central part of that problem. When orders remain encrypted until their position is final, privileged access to those protected orders no longer lets a builder read them and trade ahead. That reduces the value of early access and the incentive to concentrate it. This is a reason to build encryption into the protocol: protecting users and preserving decentralization are connected engineering goals.
Our direction combines on-chain routing with threshold encryption. Efficient computation and state access allow the routing decision to move closer to execution. Encryption protects the instruction during ordering. This is the path toward better execution for users.
The intended sequence is straightforward:
The user submits an encrypted instruction.
The network finalizes the order of encrypted transactions.
Enough decryption participants release shares to reveal the contents.
Applications execute the instructions in the fixed order, using the state available at that point.
Finalization here fixes the transaction order. Execution results follow decryption. The protection depends on the threshold security assumptions and on participants releasing shares at the correct time. Category Labs' BTX research on batched threshold encryption provides important tools for this design.
This changes what the user needs to know. They can express an outcome and its limits, while the application handles routing and the protocol protects submission.
Fair execution should be a property of the infrastructure. It is unreasonable to expect every user to become an expert in avoiding extraction before they can trade safely. If we want people to bring their financial lives on chain, we owe them a system designed to protect their interests.
More concurrent work and less control by one proposer
The next performance targets are 100 ms blocks and 200 ms finality, while preserving the decentralized, large, geographically-disparate validator set. Delivering this will continue to push Monad's best-in-class UX, especially for applications like trading that need frequent block times.
Deterministic RaptorCast (MIP-10) is the first step in this direction. It lets validators vote before they have received and decoded a full block, by making the block encoding fixed and placing all of its chunks under a single cryptographic commitment. A validator can verify one chunk against that commitment and vote, then check the complete encoding once enough chunks arrive. Consensus can therefore advance while block data is still moving through the network, removing full-block propagation from the critical path and allowing more work to happen concurrently.
Cadence takes concurrency further. Each slot has its own consensus instance, so later slots can proceed before earlier ones finish or propagate. Multiple validators propose content for each slot. The network agrees which proposals to include and combines their transactions under a fixed rule.
Cadence also limits each proposer's power over access to a block. Its short-term censorship resistance guarantees that, after the network stabilizes and its recovery period passes, every honest proposer's on-time proposal is included in that slot, subject to the protocol's fault assumptions. Another proposer cannot suppress that contribution or move it to a later slot. Proposal encryption keeps honest proposals hidden until the slot deadline, so competitors cannot read them and construct a new proposal for the same slot in response.
Cadence and BTX work together - protecting users both when they submit a transaction and when proposers compete to include it.
Privacy and shared settlement
Many real-world businesses need confidentiality. Banks must protect customer records, and trading firms must protect positions. We cannot ask them to expose their books as the price of participation.
Private chains provide confidentiality at the cost of isolation. Assets and liquidity become fragmented across separate networks, with bridges and custom integrations needed to reach outside markets. If every institution builds its own closed system, we reproduce the fragmentation of our current financial system.
We can deliver the best of both worlds: private environments with access to shared liquidity. For example, two banks could maintain separate private environments that keep their own customer records confidential, but settle a payment and obtain currency conversion through a common market. Auditors could receive the access they need without making those records public.
Best-in-class primitives for app-specific sequencing, indexing, and authentication
Applications also need more control over how user actions reach settlement. An exchange may need to process cancellations before new orders. A game may need to acknowledge an action before the next block. These choices affect the product, yet teams often have to build custom infrastructure to support them.
Alongside native rollups, we want better support for application-specific sequencing and fast responses, with settlement on Monad. Developers should be able to choose ordering rules that serve their users. Users still need clear guarantees about when a response becomes final and how to submit transactions if the application operator stops responding.
Those connections need clear rules for asset transfers, withdrawals, and recovery if an operator fails. Sharing a settlement layer does not automatically make every interaction instant or atomic. The goal is to make private and public activity work together reliably.
Usability matters at the application boundary too. MIP-16 proposes consistent query methods for blocks, transactions, logs, traces, and transfers. Mera, already available in preview, lets developers create accounts from passkeys and support signing sessions. These reduce different kinds of friction: obtaining data and helping a user access an account.
Building best-in-class primitives for authentication, indexing, and application-specific sequencing allows developers to spend more time on their core product.
Quantum security
A financial system must protect assets over many years. Its security cannot depend on cryptographic assumptions remaining suitable forever. Waiting for a crisis before giving users a migration path would be a failure of engineering.
We therefore want a quantum-resistant Monad. The work must cover and the cryptography that protects the network, with a practical transition for existing accounts and applications.
The highest priority is protecting user funds. The flexible and upgradeable account authorization plan addresses this, allowing a graceful transition to PQ signature schemes without disrupting DeFi or existing address books.
After that, the other priorities are:
protecting consensus by implementing a post-quantum version of Cadence
protecting validator authentication
adding precompiles for PQ signature verification
migrating the zkVM proof infrastructure
protecting the signature schemes in BTX
The endgame architecture is within reach
Bringing all of the above capabilities together will deliver a world computer worthy of housing the world's money. Solving these problems - caring for user, developer, and business happiness - has cascading effects: more efficient markets, better pricing, better asset selection, more capital, more opportunity. Users get access to the best opportunities, unlocked by a borderless financial system with composable money legos.
This revolution can only happen on a decentralized, neutral system. A wallet, exchange, or financial institution should be able to build without depending on infrastructure whose rules favor a competitor. This is central to the general-purpose network we set out to build.
If we accomplish these goals, we get a new financial system that meets the expectations of users and the needs of businesses. People can control their money, submit an order without inviting someone to trade ahead of them, and receive a fast result. Businesses can protect customer information, meet their disclosure obligations, and transact with the wider economy. Developers can compete to serve both on infrastructure that is open to everyone.
The result is a system that finally satisfies the expectations that users and businesses have in the rest of finance, and that is future-proof.
You can also read this post on my blog.(Historical earnings: For fiscal 2026 Q3 (period ended 2026-05-28), Micron Technology reported basic EPS of 41.97, diluted EPS of 41.4, and net income of USD 47.268 billion. In the prior-year period (ended 2025-05-29), basic EPS was 4.79, diluted EPS was 4.75, and net income was USD 5.338 billion. For fiscal 2026 Q2 (period ended 2026-02-26), basic EPS was 16.91, diluted EPS was 16.68, and net income was USD 19.025 billion; in the prior-year period (ended 2025-02-27), basic EPS was 3.1, diluted EPS was 3.08, and net income was USD 3.453 billion. Consensus expectations: The next upcoming consensus is for fiscal 2026 Q4, with an EPS estimate of 32.5625 and no revenue estimate available.)