V God发布 Diamond iO 混淆方案，大幅压缩 iO 算力开销，隐私密码学或迎新突破
On July 28, Ethereum co-founder Vitalik Buterin published the technical article “Obfuscation (Part 2): Diamond iO,” focusing on a new indistinguishability obfuscation (iO) cryptography scheme called Diamond iO. The scheme is designed to address the core pain point of traditional iO suffering from excessively high computational overhead.
Diamond iO enables programs to be encrypted so that, after receiving plaintext inputs, they can run normally and output results, while fully hiding the program’s internal logic and built-in keys. This provides important technical support for sensitive computing scenarios.
The computation time of traditional mainstream conservative iO protocols reaches a “cosmic-level,” and in theory they cannot be practically deployed. In contrast, Diamond iO improves the BGG+14 attribute encryption (ABE) architecture, successfully compressing the running overhead to a “planetary-level.”
Although this cryptography scheme is still not ready for commercial use, with only a small amount of optimization it is expected to be adapted to some application scenarios, bringing new possibilities for the practical deployment of program obfuscation technology.
The scheme’s core architecture is fully homomorphic encryption (FHE) nested with ABE. Efficiency improvements are achieved through three main mechanisms: designing dedicated conditional decryption logic, modifying function outputs to produce pseudorandom results, and using a tensor fusion mechanism so users can generate input encodings independently.
Compared with traditional multi-layer nested constructions, the Diamond iO architecture is more concise, with computational complexity equivalent to ordinary function encryption, thereby significantly reducing redundant computations.
From a security perspective, the scheme relies on two new cryptographic assumptions—All-product LWE and Evasive LWE—that have not yet been widely verified, so there is theoretical controversy. The industry still needs extensive security reasoning and validation to establish reliability;
From an engineering deployment perspective, the scheme only fits low-depth circuits. Hash operations are the main performance bottleneck, and it has drawbacks such as oversized parameter scales and accumulated noise. Vitalik outlines optimization paths such as replacing efficient PRGs and streamlining the FHE nesting layers.
Overall, once Diamond iO is deployed, it could support scenarios such as encrypted private key custody, privacy-preserving smart contracts, and untrusted encryption services, providing a new foundational tool for Web3 privacy infrastructure. However, it still has a long development cycle before large-scale real-world deployment.
#Vitalik #Web3隐私
On July 28, Ethereum co-founder Vitalik Buterin published the technical article “Obfuscation (Part 2): Diamond iO,” focusing on a new indistinguishability obfuscation (iO) cryptography scheme called Diamond iO. The scheme is designed to address the core pain point of traditional iO suffering from excessively high computational overhead.
Diamond iO enables programs to be encrypted so that, after receiving plaintext inputs, they can run normally and output results, while fully hiding the program’s internal logic and built-in keys. This provides important technical support for sensitive computing scenarios.
The computation time of traditional mainstream conservative iO protocols reaches a “cosmic-level,” and in theory they cannot be practically deployed. In contrast, Diamond iO improves the BGG+14 attribute encryption (ABE) architecture, successfully compressing the running overhead to a “planetary-level.”
Although this cryptography scheme is still not ready for commercial use, with only a small amount of optimization it is expected to be adapted to some application scenarios, bringing new possibilities for the practical deployment of program obfuscation technology.
The scheme’s core architecture is fully homomorphic encryption (FHE) nested with ABE. Efficiency improvements are achieved through three main mechanisms: designing dedicated conditional decryption logic, modifying function outputs to produce pseudorandom results, and using a tensor fusion mechanism so users can generate input encodings independently.
Compared with traditional multi-layer nested constructions, the Diamond iO architecture is more concise, with computational complexity equivalent to ordinary function encryption, thereby significantly reducing redundant computations.
From a security perspective, the scheme relies on two new cryptographic assumptions—All-product LWE and Evasive LWE—that have not yet been widely verified, so there is theoretical controversy. The industry still needs extensive security reasoning and validation to establish reliability;
From an engineering deployment perspective, the scheme only fits low-depth circuits. Hash operations are the main performance bottleneck, and it has drawbacks such as oversized parameter scales and accumulated noise. Vitalik outlines optimization paths such as replacing efficient PRGs and streamlining the FHE nesting layers.
Overall, once Diamond iO is deployed, it could support scenarios such as encrypted private key custody, privacy-preserving smart contracts, and untrusted encryption services, providing a new foundational tool for Web3 privacy infrastructure. However, it still has a long development cycle before large-scale real-world deployment.
#Vitalik #Web3隐私