The "New Story" of Decentralized Computing: Will Quilibrium Be the Next ICP?

1. Key Points of the Report

1.1 Core Investment Logic

• Quilibrium seeks to find a "balance" between the computing power of traditional internet and the decentralization of blockchain, and has designed a unique decentralized cloud computing architecture for this purpose.
• Quilibrium has built a database-based operating system that is closer to traditional software in terms of development experience, which may attract more traditional software developers and facilitate current Web3 developers in building more complex crypto applications.
• Quilibrium's design emphasizes security and privacy, making it highly attractive for enterprises that do not wish to expose sensitive data but want to utilize encryption technology; for individuals, the initial breakout of Farcaster also proves the long-term potential of decentralized applications in acquiring users and generating revenue.
• Founder and CEO Cassie Heart is a former senior engineer at Coinbase and a Farcaster developer, and the team possesses rich experience, stable delivery capabilities, and a distinct personality.

1.2 Major Risks

• The project is in a very early stage, with the mainnet yet to be launched, and the project's complexity is high, with verification of technical feasibility and market demand still incomplete.
• In the short term, it may face competition from the more well-known Arweave AO in terms of user perception and developer engagement.
• There is no fixed token model, and the token release rate may be unstable, which adds a certain level of risk for investors.

1.3 Valuation

As Quilibrium is still in a very early stage, we cannot yet derive an accurate valuation for the project. However, from the perspective of circulating market capitalization and fully diluted market capitalization, Quilibrium's current market cap has certain attractiveness compared to other market players with overlapping concepts.

2. Business Analysis

Quilibrium positions itself as a "decentralized internet layer protocol that provides the convenience of cloud computing without sacrificing privacy or scalability" and "a decentralized PaaS solution." This section will elaborate on Quilibrium's business around the following questions.

• What are the problems with traditional internet cloud computing?
• Why do we need (another) decentralized computer?
• What makes Quilibrium special compared to current mainstream blockchain designs?

Source: Cassie Heart's Farcaster account

2.1 Business Positioning

2.1.1 Starting with Computing

Whether in Web2 or Web3, "computing" is a crucial concept and the driving force behind application development, execution, and scaling.

In traditional internet architecture, computing tasks are typically completed by centralized servers. The emergence of cloud computing has improved the scalability, accessibility, and cost-efficiency of computing, gradually replacing traditional computing as the mainstream.

From the service content perspective, the cloud service models provided by large cloud service providers can generally be divided into three categories: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS), corresponding to different needs and capabilities of entities, providing varying levels of control over resources. End users are generally more familiar with SaaS. PaaS and IaaS are primarily aimed at developers. Source: Lydia @ Mint Ventures Source: S2 Lab, Lydia @ Mint Ventures

In mainstream blockchains like Ethereum, computing is usually performed by decentralized nodes. This method does not rely on centrally controlled servers; each node executes computing tasks locally and ensures data correctness and consistency through consensus mechanisms, but the capabilities and processing speeds of decentralized computing typically cannot compare with traditional cloud services.

Quilibrium seeks to find a "balance" between the computing power and scalability of traditional internet and the decentralization of blockchain, opening up new possibilities for application development. Source: Cassie Heart's live stream recording

2.1.2 Centralization Issues of Computer Systems

For most end users, the centralization issue of computers is not easy to perceive. This is because what end users directly face is mostly hardware-level computer systems. Our PCs, mobile phones, and other devices are scattered around the world and operate independently under personal control. This distributed physical presence means that computer systems are not necessarily centralized at the hardware level.

In contrast to relatively decentralized hardware, existing computer systems are significantly more centralized at the network architecture and cloud computing service levels—Amazon AWS, Microsoft Azure, and Google Cloud accounted for over 67% of the cloud service market share in Q1 2024, creating a significant gap with later entrants. Source: Synergy Research Group

Moreover, as the "water sellers" of the AI wave, the trend of strong cloud service providers getting stronger seems to continue. Microsoft Azure, as the exclusive cloud service provider for OpenAI, has seen its performance growth accelerate over the past year, reversing previous declines. In Microsoft's Q3 FY2024 (i.e., Q1 of the 2024 calendar year) earnings report, revenue from Azure and other cloud services grew by 31%, exceeding the market's expected growth of 28.6%. Source: Microsoft, Lydia @ Mint Ventures

In addition to market competition considerations, the privacy and security issues brought about by centralized computer systems are also receiving increasing attention—every outage of a few large cloud service providers can have widespread impacts. Data shows that from 2010 to 2019, AWS experienced a total of 22 outages, averaging 2.4 outages per year. In addition to affecting Amazon's own e-commerce business, the network services of companies using AWS, such as Robinhood, Disney, Netflix, and Nintendo, were also massively interrupted.

2.1.3 The Proposal of Decentralized Computers

Against this backdrop, the necessity of decentralized computers has been repeatedly proposed. In recent years, centralized cloud service providers have increasingly adopted distributed architectures to avoid single points of failure by replicating data and services across multiple locations and enhancing performance through edge storage. The narrative of decentralized computing has gradually shifted to focus on data security, privacy, scalability, and cost-effectiveness.

Let’s first clarify several concepts of decentralized computers proposed by different projects, which share the common feature of aiming to build a global distributed computing platform through decentralized data storage and processing to support the development of decentralized applications.

• World Computer: Generally refers to Ethereum, providing a global smart contract execution environment, with its core function being decentralized computing and the global unified execution of smart contracts.
• Internet Computer: Generally refers to the ICP developed by the Dfinity Foundation, aiming to extend the functionality of the internet and enable decentralized applications to run directly on the internet.
• Hyper Parallel Computer: Generally refers to the AO protocol proposed by Arweave, which is a distributed computing system running on the Arweave network, characterized by high parallelism and high fault tolerance.

It is worth noting that ICP, AO, and Quilibrium are not traditional blockchains in the conventional sense. They do not rely on a linear block arrangement structure but maintain core principles of blockchain such as decentralization and data immutability, and can be seen as a natural extension of blockchain technology. Although ICP has yet to realize its grand vision, the emergence of AO and Quilibrium indeed brings new possibilities that could influence the future of Web3.

The table below compares the technical characteristics and application directions of the three, aiming to help readers understand "Will Quilibrium repeat the mistakes of ICP," and how Quilibrium differs from AO, which is referred to as the "Ethereum killer," as another frontier solution for decentralized computing.

2.2 Consensus Mechanism

In traditional blockchains, the consensus mechanism is at a relatively abstract and core level, defining how the network reaches consensus, processes and verifies transactions, and performs other operations. Different choices of consensus mechanisms will affect the network's security, speed, scalability, and degree of decentralization.

Quilibrium's consensus mechanism is called "Proof of Meaningful Work (PoMW)," where miners are required to complete tasks that have actual significance for the network, such as data storage, data retrieval, and network maintenance. The design of the PoMW consensus mechanism integrates multiple fields such as cryptography, multiparty computation, distributed systems, database architecture, and graph theory, aiming to reduce reliance on a single resource (such as energy or capital), ensure the degree of decentralization of the network, and maintain security and scalability as the network scales.

The incentive mechanism is key to ensuring the smooth operation of the consensus mechanism. Quilibrium's incentive distribution is not static but dynamically adjusted based on network conditions to ensure that incentives match demand. Quilibrium also introduces a multi-proof mechanism, allowing a node to verify multiple data segments, thus maintaining network operation even when nodes and core resources are insufficient.

We can understand the final earnings of miners with a simplified formula, where the unit reward is dynamically adjusted based on the network scale.

Earnings = Score × Unit Reward

The score is calculated based on various factors, with the specific formula as follows:

Where the parameters are defined as follows:

1. Time in Mesh for Topic: The longer the participation time and the higher the stability, the higher the score.
2. First Message Deliveries for Topic: The more times messages are delivered for the first time, the higher the score.
3. Mesh Message Delivery Rate/Failures for Topic: Nodes with high delivery rates and low failure rates score higher.
4. Invalid Messages for Topic: The fewer invalid messages delivered, the higher the score.

The weighted sum of the above four parameters will have a topic cap (TC), which serves to limit this value within a certain range to avoid unfair scoring caused by excessively large parameters.

5. Application-Specific Score: A score defined by specific applications.
6. IP Collocation Factor: The fewer nodes from the same IP address, the higher the score.

Source: Quilibrium Dashboard

Quilibrium currently has over 60,000 running nodes, and the actual earnings of nodes in operation may fluctuate based on the different parameter weights between each version. After version v1.4.19, miner earnings can be viewed in real-time, but they can only be claimed after the mainnet goes live.

2.3 Network Architecture

Quilibrium's core business is a decentralized PaaS solution, and its network architecture mainly consists of communication, storage, data querying and management, and operating systems. This section will focus on the differences in its design compared to current mainstream blockchains. Readers interested in technical details and implementation methods can refer to the official documentation and white paper.

2.3.1 Communication

As the foundational infrastructure of the network, Quilibrium's communication consists of four parts.

a. Key Generation

Quilibrium proposes a graph theory-based PCAS (Planted Clique Addressing Scheme) key generation method. Similar to traditional blockchain technology, PCAS also uses asymmetric encryption—each user has a public key and a private key, where the public key can be shared for encrypting information or verifying signatures, while the private key is kept secret for decrypting information or generating signatures. The differences mainly lie in the key generation methods, representations, and application directions (see the table below).

b. End-to-End Encryption

End-to-end encryption (E2EE) is a key component to ensure secure communication between nodes, where only the communicating parties can see the plaintext data, and even the systems or intermediaries that help transmit the information cannot read the content.

Quilibrium employs a method called Triple-Ratchet for end-to-end encryption, which provides higher security compared to traditional ECDH schemes. Specifically, traditional schemes typically use a single static key or periodically update keys, while the Triple-Ratchet protocol updates keys after each communication, achieving forward secrecy, backward secrecy, resilience, replay protection, and unordered message delivery. This scheme is particularly suitable for group communication but is relatively more complex and computationally expensive.

c. Mix Network Routing

Mix networks (Mixnets) act as a black box that receives messages from senders and transmits them to receivers, while external attackers cannot associate senders and receivers even if they can access information outside the black box.

Quilibrium utilizes RPM (Random Permutation Matrix) technology, providing a mix network architecture that is structurally complex and difficult for external and internal attackers to crack, offering advantages in anonymity, security, and scalability.

d. Peer-to-Peer Communication

GossipSub is a peer-to-peer messaging protocol based on a publish/subscribe model, widely used in blockchain technology and decentralized applications (DApps). Quilibrium's BlossomSub protocol is an extension and improvement of the traditional GossipSub protocol, aimed at enhancing privacy protection, increasing resistance to Sybil attacks, and optimizing network performance.

2.3.2 Storage

Most traditional blockchains use cryptographic hash functions as basic tools for data integrity verification and rely on consensus mechanisms to ensure network consistency. This mechanism has two main limitations:

• It typically does not include verification of storage time, lacking a direct mechanism to defend against time-based or computational power-based attacks.
• Storage and consensus mechanisms are usually quite separate, which may lead to data synchronization and consistency issues.

Quilibrium's storage solution employs a Verifiable Delay Function (VDF) design, creating a time-dependent chain structure that integrates storage and consensus mechanisms. Combining the diagram below, we can summarize several characteristics of this solution:

• Input Processing: By using hash functions such as SHA256 and SHAKE128 to process inputs, any slight change in data will lead to a significant difference in hash values, making data harder to tamper with and easier to verify.
• Delay Guarantee: The computation process is deliberately set to be time-consuming. Computing tasks must be executed in order, with each step relying on the results of the previous step, and cannot be accelerated by increasing computational resources, ensuring that outputs are derived from continuous and deterministic time-based computations. Since the generation process cannot be parallelized, any attempt to recompute or alter already published VDF results would require a considerable amount of time, providing network participants with sufficient time to detect and respond.
• Fast Verification: The time required to verify a VDF result is much less than that needed to generate that result, typically requiring only some mathematical checks on the final result or using auxiliary data to confirm the validity of the result.

Source: Quilibrium White Paper

This time-proof-based chain structure does not rely on the generation of blocks in traditional blockchains, theoretically reducing MEV attacks and front-running phenomena.

2.3.3 Data Querying and Management

Most traditional blockchains use simple key-value storage or Merkle Trees to manage data, which typically limits their ability to express complex relationships and support advanced queries. Moreover, most current blockchain systems do not provide built-in privacy protection mechanisms when nodes execute queries, which is the background for the emergence of privacy-enhancing technologies like zero-knowledge proofs.

Quilibrium proposes an "Oblivious Hypergraph" architecture, combining hypergraph structures and Oblivious Transfer technology, which can support complex querying capabilities while maintaining data privacy. Specifically:

• Hypergraph Structure: Allows edges to connect multiple vertices, enhancing the ability to express complex relationships. This structure can directly map various database models, allowing any type of data relationship to be expressed and queried on the hypergraph.
• Oblivious Transfer Technology: Even the nodes processing the data cannot understand the specific content of the data being accessed, enhancing privacy protection during the data querying process.

2.3.4 Operating System

The operating system is not a native concept of blockchain. Most traditional blockchains primarily focus on consensus mechanisms and data immutability, usually not providing complex operating system-level functionalities. For example, while Ethereum supports smart contracts, its operating system functionalities are relatively simple, mainly limited to transaction processing and state management.

Quilibrium has designed an operating system based on its hypergraph database and implemented common operating system primitives, such as file systems, schedulers, IPC-like mechanisms, message queues, and control key management. This design, which builds the operating system directly on the database, can support the development of complex decentralized applications. Source: Quilibrium White Paper

2.4 Programming Language

Quilibrium's development primarily uses Go as the main programming language, along with Rust and JavaScript. The advantages of Go lie in its ability to handle concurrent tasks, concise syntax, and an active developer community. According to the Tiobe programming language rankings, Go has seen a significant rise in rankings in recent years, ranking 7th in the latest June list. Other blockchain projects that also use Go for underlying development include Ethereum, Polygon, and Cosmos. Source: Quilibrium Source: Tiobe

3. Project Situation

3.1 Project History and Roadmap

Quilibrium's white paper was released in December 2022, and its roadmap is roughly divided into three phases: Dusk, Equinox, and Event Horizon.

Quilibrium is still in a very early stage, with the team conducting network updates and iterations every two weeks. The latest version is v1.4.20. Since the team has removed the 1.5 phase from the roadmap, the network will directly upgrade from version 1.4 to version 2.0. Version 2.0, which is the mainnet, marks the end of the Dusk phase and is expected to be officially launched in late July, at which point bridging for $QUIL will be allowed.

According to the tentative plan, the Equinox and Event Horizon phases will support more advanced applications such as streaming and AI/ML model training.

3.2 Team and Financing

The founder/CEO of Quilibrium is Cassie Heart. Before founding Quilibrium, she was a senior software engineer at Coinbase, with over 12 years of experience in software development and blockchain.

Due to Cassie's opposition to centralized social media platforms, she and Quilibrium's project account are mainly active on Farcaster. Cassie's Farcaster account has over 310,000 followers, including Ethereum founder Vitalik. Cassie is also a developer of Farcaster.

From Quilibrium's developer data dashboard, the development of the Quilibrium project began in April 2023 and has been stable since then. There are a total of 24 developers displayed, with Cassie Heart (Cassandra Heart) being the main contributor. Source: Quilibrium

Quilibrium's team has not disclosed its financing history and investment institutions.

3.3 Token Model Analysis

$QUIL is Quilibrium's native token, launched through a 100% fair launch model, with all token outputs coming from node operations. The team operates a small number of nodes, but their token ownership accounts for less than 1%.

$QUIL does not have a fixed token model, and the total token supply is unlimited, but it will be dynamically adjusted based on the speed of network adoption—when the network scale increases, more tokens will be released as node incentives; if the growth slows, the token release rate will decrease accordingly.

The table below shows the predictions made by the team and community members regarding the token release schedule. The current circulating supply is 340 million, and the expected final supply will converge around 2 billion, with the specific release situation depending on ecological development. Source: @petejcrypto

3.4 Risks

Potential risk points for Quilibrium at this stage include:

• The project is in a very early stage, with the mainnet yet to be launched, and the project's complexity is high, with verification of technical feasibility and market demand still incomplete.
• In the short term, it may face competition from the more well-known Arweave AO in terms of user perception and developer engagement.
• There is no fixed token model, and the token release rate may be unstable, which adds a certain level of risk for investors.

4. Valuation

The valuation of quasi-public chain infrastructure is inherently a very complex process, involving multiple dimensions such as TVL, on-chain active addresses, number of dApps, and developer community. Since Quilibrium is still in a very early stage, and the token $AO of Arweave AO has not yet been opened for trading, we cannot derive an accurate valuation for the project at this time.

We list the circulating market capitalization and fully diluted market capitalization of projects that have certain conceptual overlaps with Quilibrium (data as of June 23, 2024) for reference. Source: CoinGecko, data as of June 23, 2024

5. References and Acknowledgments

The writing of this article owes thanks to the reviews and opinions of Hai Ge (@PleaseCallMeWhy), Lan Ge, and Connor.

• https://quilibrium.com/quilibrium.pdf
• https://paragraph.xyz/@quilibrium.com
• https://dashboard.quilibrium.com/
• https://www.youtube.com/watch?v=Ye677-FkgXE\&ab_channel=CassandraHeart
• https://dune.com/cincauhangus/quilibrium
• https://source.quilibrium.com/quilibrium/ceremonyclient/-/graphs/main?ref_type=heads
• https://www.tiobe.com/tiobe-index/
• https://www.blocmates.com/meal-deal-research-reports/quilibrium-crypto-not-blockchain-long-live-the-internet
• https://www.statista.com/chart/18819/worldwide-market-share-of-leading-cloud-infrastructure-service-providers/
• https://s2-labs.com/admin-tutorials/cloud-service-models/
• https://medium.com/@permadao/%E5%8E%BB%E4%B8%AD%E5%BF%83%E5%8C%96%E4%BA%91%E6%9C%8D%E5%8A%A1%E8%BF%9B%E5%8C%96%E5%8F%B2-%E4%BB%8E-dfinity-ic-%E5%88%B0-arweave-ao-839b09b4f3ff
• https://www.microsoft.com/en-us/investor/earnings/FY-2024-Q3/press-release-webcast
• https://x.com/perma_daoCN/status/1798565157435830416
• https://x.com/Pow2wer/status/1802455254065402106