The Brevis team is proud to introduce Pico—a modular and efficient zero-knowledge virtual machine (zkVM). Pico allows developers to build zkVM flexibly like assembling Lego bricks, optimizing performance and user experience based on the computational needs of the application. Developers can freely choose from a rich set of built-in options, or even fully customize the proof backend and virtual machine instances to build a dedicated computational workflow that meets their own requirements.

Pico redefines the "Glue-and-Coprocessor" architecture, not only supporting underlying coprocessors (such as using precompiled modules to accelerate specific VM instruction operations), but also natively integrating Brevis' on-chain data zkCoprocessor, which can boost the performance of applications utilizing blockchain historical data by up to 32 times.

Although Pico has not yet released a complete GPU-accelerated cluster solution, it has already achieved world-class performance on the CPU side. Compared to other industry zkVMs such as RISC0, SP1, and OpenVM, Pico's CPU execution speed is 70% to 155% faster, setting a new performance benchmark.

With the release of Pico v1.0, we introduce the world's first zkVM that supports customizable computational architectures, allowing developers to flexibly adjust the following key parameters:

Optional proof backends: Support STARK on KoalaBear and BabyBear, as well as CircleSTARK on Mersenne 31.

Flexible proof workflows: Optimize security, scalability, and proof generation efficiency to adapt to different application requirements.

Access to on-chain historical data: Through the built-in on-chain data zkCoprocessor, developers can freely query and compute historical blockchain data to build dApps, achieving optimal computational performance and programming flexibility.

Pico is compatible with the RISC-V instruction set and supports the Rust development toolchain. Due to its modular architecture, Pico can be continuously upgraded in the future, easily adapting to the latest ZK theory research and innovations, whether for next-generation zk applications or frontier explorations in the ZK field, Pico can provide developers with a stable and powerful computational foundation.

Please visit the Pico Developer Guide, explore the GitHub code repository, and join the Telegram or Discord discussion group to stay up-to-date and participate in the contribution, together we will drive the future of ZK computing.

Why build Pico?

Brevis' first product, the on-chain data zkCoprocessor, has been widely used in the DeFi field. Many partners have already deployed innovative features based on Brevis on the mainnet, including Kwenta, Usual, Algebra Labs, JoJo Exchange, and Trusta. Additionally, top protocols and applications such as PancakeSwap, Celer, Gamma, Quickswap, Frax, Mask Network, Kernel, BeraBorrow, Thena, Kim Protocol, 0G, Bedrock, Mellow Finance, ZettaBlock, and Hemera are also developing next-generation products and features based on Brevis.

And Mendi Finance are among the many top protocols and applications that are developing next-generation products and features based on Brevis.

However, in the course of collaborating with these pioneering teams, we have found that the actual application demands for zk computing are becoming increasingly diverse, and they are facing the following core challenges:

1. Adapting to different application requirements

The business logic of different projects varies, and their performance requirements are also vastly different. Traditional "one-size-fits-all" zkVMs or fixed ZK circuit solutions cannot meet these diverse needs. If there is a lack of flexibility in customizing the proof workflow, or the inability to integrate custom circuits (whether at the instruction code level or the application level), the scalability of the zkVM will be limited.

2. Adopting the latest ZK technologies

The ZK field is developing rapidly, with new proof backends, frameworks, and cryptographic breakthroughs emerging constantly. Many existing zk solutions, due to their closed architectures, find it difficult to quickly integrate these innovative technologies, resulting in outdated proofs, poor performance, high computational costs, and limited optimization potential.

Pico's Modular Solution: Truly Flexible and Scalable

To address these challenges, Pico adopts a modular architecture and provides:

· Support for multiple proof backends: Easily switch or upgrade to the latest proof backends to ensure computational performance and compatibility.

· Customizable proof workflows: Developers can freely customize the proof generation pipeline to meet the specific needs of their applications.

· Extensible coprocessor integration: Support for building or integrating dedicated coprocessors, without being limited by a fixed zkVM framework.

Innovative "Glue-and-Coprocessor" Architecture: Breaking the Bottleneck of Traditional zkVMs

Pico adopts the "Glue-and-Coprocessor" architecture, combining efficient specialized circuits (Coprocessors) and a general-purpose zkVM (Glue), striking a balance between performance, generality, and scalability.

Coprocessors: Optimize compute-intensive tasks (such as arithmetic operations, cryptographic computations, machine learning, etc.) to improve ZK proof efficiency.

General-purpose zkVM (Glue): Responsible for managing the overall proof and verification workflow, ensuring that all logical computations can be performed securely.

What are the advantages of this architecture?

Compared to traditional zkVMs, Pico, by combining the generality and the computational power of specialized circuits, generates proofs faster and has higher programming flexibility.

Precompiles are a common type of coprocessor, extending the RISC-V instruction set to accelerate underlying operations such as hash calculations and signature verifications. Pico supports developers to customize precompile modules based on their needs, while also providing out-of-the-box optimization solutions.

However, precompiles alone are not enough to solve the performance requirements of all applications. For example, if a developer wants to prove that a trader has completed 10,000 Uniswap transactions totaling $50 million in the past 30 days, using only a zkVM to perform the calculation would require writing a Merkle tree inclusion proof program and an RLP parsing program, resulting in high execution costs.

How to break through this limitation?

Pico natively integrates Brevis' on-chain data zkCoprocessor as an application-level coprocessor, allowing developers to efficiently access and compute on-chain historical data, achieving a 32-fold performance improvement and a 67% cost reduction.

By combining coprocessors and a general-purpose zkVM (Glue), Pico provides developers with a powerful and flexible tool to balance performance, programmability, and adaptability in ZK-driven applications.

Flexible Proof Backends and Customizable Computational Workflows

1. Flexible Support for Multiple Proof Backends

Pico is compatible with various zero-knowledge proof systems, including:

STARK (KoalaBear, BabyBear)

CircleSTARK (Mersenne 31)

For example, the Poseidon2 hash function is widely used in zkVM recursive proofs. Under the same STARK system, KoalaBear's proof efficiency is far superior to BabyBear's, and the performance can be greatly improved by simply replacing the proof backend, without the need to modify the computation logic.

2. Customizable Proof Workflow

Pico allows developers to freely adjust the proof generation process to optimize scalability, cost, and latency.

· Instance-level Optimization: Developers can customize the computation process (including proof backend, computing chips, memory management, etc.) for each VM instance.

· ProverChain Proof Chain: Optimize computational efficiency through a modular process of RISCV → CONVERT → COMBINE → COMPRESS → EMBED → ONCHAIN.

· Optional Decentralized Verification: Developers can choose whether to perform on-chain verification on EVM to balance performance and decentralization.

Refreshing Industry Performance Records: Setting a New Benchmark for zkVM

In the latest performance benchmark tests, Pico has comprehensively surpassed existing zkVM solutions and demonstrated remarkable performance improvements in CPU computing environments. We conducted comparative tests on RISC0, SP1, and OpenVM, covering the following core computational tasks:

· Fibonacci Calculation

· Tendermint Block Consensus

· Proof of Ethereum Reth Block #17106222

All tests were conducted on AWS r7a.48xlarge instances (192 CPU cores, 1.5TB RAM) to ensure consistency in the computing environment. The results show:

· Pico's running speed in all tasks is up to 155% faster than the second-fastest solution, setting a new record for zkVM computing performance!

· Pico's CPU computing capabilities are significantly ahead, making it particularly suitable for applications that require efficient computation.

GPU-accelerated Version Coming Soon

Although the current data is based on CPU testing, Pico is developing a GPU-accelerated version, which is expected to be released in the coming months. The GPU-accelerated version of Pico will further enhance its zk computing capabilities, providing even greater throughput and computational efficiency. We will release a comprehensive GPU performance test report in the future.

Standing on the Shoulders of Giants

Pico has drawn inspiration from the following projects, each of which represents a frontier advancement in zero-knowledge proof systems.

By building upon these innovations, Pico provides a modular and high-performance zkVM:

Plonky3: Pico's proof backend is based on Plonky3, and by extending its modular features to the zkVM layer, it allows developers to flexibly choose the most suitable proof domain and proof system for their applications.

SP1: Pico has gained important insights from SP1's chip design and constraint system, including the design and constraints of the recursion compiler and Precompiles.

Valida: Pico's implementation of cross-table lookups was inspired by Valida's pioneering work in this field.

RISC0: Pico's Rust toolchain is directly based on the toolchain originally developed by RISC0.

Join the Pico Developer Community

Brevis has always believed that the future of zero-knowledge technology lies in collaboration and innovation. Pico is not just a zkVM, but a development platform that empowers the construction of the next generation of zk applications.

· Read the Pico Developer Documentation: Pico Docs

· Explore the GitHub Repository: GitHub Repo

· Join the Community Discussions: Telegram or Discord

Let's work together to expand the boundaries of zero-knowledge computing and build a more intelligent, trustless, and decentralized world!

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Twitter Official Account: https://twitter.com/BlockBeatsAsia