Anaxi Labs and Carnegie Mellon University’s CyLab, a network security and privacy research institute, announced the launch of a revolutionary cryptography compiler framework, addressing a long-standing industry challenge—building scalable zero-knowledge applications requires trade-offs at the fundamental level. Prior to this, the three features of scalability, encryption security, and Decentralization were considered impossible to achieve simultaneously, which has been a barrier to large-scale applications—until now.

Cryptography Breakthrough Without Sacrifice

Blockchain such as Ethereum is hailed as the future infrastructure of Decentralization, and zero-knowledge (ZK) technology is believed to enhance the security and scalability of Ethereum to over 120 transactions per second. However, the reality is quite different. Developing Zero-Knowledge Proof is both complex and time-consuming, requiring dozens of developers to invest thousands of hours. To optimize proof generation speed, manual design of the protocol is often required, and manual coding and tens of thousands of lines of code bring significant security risks. This makes creating security-sensitive Decentralization applications complex and makes auditing and Compliance a nightmare - all of which hinder the widespread application of this technology in regulated industries such as finance, healthcare, and artificial intelligence.

The research team at Carnegie Mellon University is working with Anaxi Labs to overcome this challenge

Carnegie Mellon University’s latest paper proposes a revolutionary approach that can automatically compile and simplify high-level software into a simpler form (low-level representation) required by the underlying proof system. This process is fully automated, repeatable, and auditable, eliminating manual work, significantly improving performance, and ensuring the security of the process at the Cryptography level. The achievement analyzes high-level programs, decomposes them into small, indivisible units, and then creates low-level representations for easy input into various proof systems from each unit.

Assistant Professor Riad Wahby of the Department of Electrical and Computer Engineering at Carnegie Mellon University said, ‘Replacing the CPU with this automated way of decomposing computation into very specific blocks is a new approach, and it is the first time someone has attempted this compiler method that avoids representing the entire program. We are very excited about this.’

Start a new type of Decentralization application

This research and the framework built by Anaxi Labs based on the research will fundamentally change the industry landscape of Web3 and other fields. In the traditional and regulated financial field, maintaining auditability while improving performance makes real-time Settlement (such as instant dollar payment) between banks possible. In the healthcare field, in the context of the challenges faced by 23andMe, the secure and privacy-protecting encryption tools supported by the products being developed by Anaxi Labs can now address critical issues by ensuring individuals’ legitimate ownership of their DNA, while supporting valuable research. In the enterprise-level artificial intelligence and critical physical infrastructure fields, the need for high availability and near-zero latency Decentralization solutions (such as quick fine-tuning and inference across multiple data and computing resources) has also become a reality.

In recent times, products based on this research have provided the most effective solutions for Web3 enterprises addressing scalability, security, and decentralization trade-offs, offering new design paradigms for rollups and interoperability products.

Anaxi Labs co-founder Kate Shen said, “This research, as well as the product we are building that integrates the research findings, will have a profound impact on many important industry applications that require auditable solutions to address performance bottlenecks, such as ZK and EVM, ultimately taking us closer to realizing the vision of secure decentralization consensus with real-time settlement capabilities.”

Shen added, “We also appreciate the cross-language capability of this research, which means that various projects can benefit without modifying the code. This allows us to build an open, collaborative framework, rather than adopting the increasingly static, monolithic approach of today. This enables all developers to automatically select and combine the advantages of the latest proof systems, such as table lookups, coprocessors, and hardware acceleration, maximizing the performance benefits of each computation method.”

Anaxi Labs and CyLab: Changing the Game Rules Collaboration

Carnegie Mellon University’s CyLab has always been at the forefront of blockchain development (including zero-knowledge technology) research. CyLab’s renowned professor researchers include Bryan Parno, an important contributor to the development history of zero-knowledge technology. His laboratory has produced widely cited Nova paper series; and assistant professor Riad Wahby, whose research results have spawned new cryptographic technologies that fulfill the vision of the Ethereum Foundation (recently, Andreessen Horowitz’s encryption department a16z crypto has pioneered the Jolt zkVM implementation).

The research output of this compiler framework comes from the second research project in the symbiotic partnership established by Anaxi Labs and CyLab through the CMU Security Blockchain Program. This collaboration enables scholars at Carnegie Mellon University to work with Anaxi Labs, gaining insights and learning from their commercial deployment of blockchain research in Web3 and Web 2.0 applications. It allows them to find commercial solutions to the existing major issues of blockchain, bridging the gap between the known advantages of blockchain technology and large-scale applications. It also provides a springboard for students at Carnegie Mellon University to launch their careers in the Web3 field.

CyLab’s Senior Director of Collaborative Relationships, Michael Lisanti, stated: “The collaboration between Anaxi Labs and CyLab has enhanced the ability of Carnegie Mellon University researchers to work on projects with direct real-world applications, ensuring that their work has practical correlation and potential impact.”

To learn more about Anaxi Labs, please visit the official website.

For more information about the latest work of Anaxi Labs and CyLab, please visit here.

For more information about the collaboration between CyLab and Anaxi Labs, please visit here.

About Anaxi Labs

Anaxi Labs is a new type of R&D laboratory that bridges advanced academic theory with large-scale applications. We are committed to conducting original cutting-edge research, building enterprise-level, secure, and scalable Decentralization infrastructure, and driving the development of the next generation of Decentralization applications driven by cryptography.

We collaborate with top cryptographic researchers and world-class engineers who have successfully built and operated renowned products with hundreds of millions of users. Their expertise and experience provide solid technical support for our project. We are industry partners with top cryptographic academic institutions such as Carnegie Mellon University. Together, we are committed to changing the future of the internet by unleashing the potential of science for humanity, society, and the Earth.

About CyLab

CyLab at Carnegie Mellon University is the university’s security and privacy research institute. We bring together experts from all schools and departments, covering engineering, computer science, public policy, information systems, business, finance information Risk Management, humanities, and social sciences. Our mission is to catalyze, support, promote, and enhance collaborative research and education in security and privacy across departmental, disciplinary, and geographic boundaries to have a significant impact on research, education, public policy, and practice.

This article is from a submission and does not represent the viewpoint.

：