Beyond ZKP fully homomorphic encryption How FHE empowers blockchains

After listening to the Epicenter podcast interview with Zama, I was once again anxious about the lack of brain volume. To put it simply, FHE is a more advanced encryption technology than ZKP zero-knowledge proofs, which can achieve data privacy transmission in a wider range, such as cloud storage, gene prediction, biometrics, crypto private transactions, etc. It just so happens that VCs have also paid special attention to FHE recently, so let’s briefly talk about my understanding:

1. FHE stands for Fully homomorphic Encryption, which allows people to perform operations on specific forms of encrypted data without having to decrypt and expose data privacy. It is a technical concept in the broader context of the Internet, which has been around for a long time, and has recently attracted the attention of the Crypto community, mainly due to the integration of FHE functions into Ethereum by Fhenix Network and the launch of the fhEVM solution. Any developer can add the FHE function to the transaction flow as a pre-compile binary.

2. What kind of imagination space will this release?For example, in the game scene, the cards in the player’s hand can be operated by the platform to promote the game without being exposed, so as to promote the “fairness” of such games; In the DAO voting scenario, whales can participate in governance voting without revealing their voting numbers, but will not affect the voting results output of the protocol;

In the private transaction scenario, users can send fully encrypted transactions to the mempool without revealing their private information such as the transaction amount, and in the regulatory compliance scenario, goverment can monitor transactions involving the overseas flow of funds, and deduct assets from blacklisted addresses without viewing legitimate transactions. The imagination space is a step larger than the ZKP;

3. What is the difference between FHE and ZKP?To put it simply, ZKP solves the problem of consistent transmission in the state of data encryption, and the submitting data can prove the authenticity of the data to the receiving data under the premise of hiding the specific data, which is a point-to-point encryption scheme, especially the lightweight zk-SNARKS has been widely used in layer2 Rollup;

FHE, on the other hand, is to solve the problem of recomputation in the state of encrypted data transmission, which is a more extensive many-to-many encryption scheme. The reason why it is not so widespread is that it consumes a lot of computing power and is inefficient in the process.

4. ZAMA has established the fhEVM solution, which can allow FHE to be applied to the blockchain field, and Phenix network is similar to SDK embedding integration, modularizing FHE into a precompiled process to add more blockchain paths, but it is currently more suitable for OP-rollup, and even can set up a pipeline to load FHE functions only for specific transactions.

This means that users can choose whether to initiate transactions through FHE on layer2 to protect their privacy, which is a little higher than the nature transaction fee, but has more security and privacy. This is bound to attract some MM market makers as well as institutional users who are concerned about private trading. The requirements and scenarios have been understood, and the key lies in the integrated tradeoff of efficiency and performance as well as fees.

Personally, I believe that privacy trading will be the next refined trend track after the transaction expansion problem.

At present, the core of the scaling problem is no longer infra technology, but the necessary conditions such as regulation and market mass adoption are lacking, while privacy transactions are further subdivided into users, transactions and usage scenarios through technical infra optimization, and to some extent, there will be a wide range of applications similar to ZKP technology.

We can imagine that with the blessing of FHE technology, the transaction experience gap between DEXs and CEXs will be reduced, some layer3 privacy application chains will have a more closed-loop and secure transaction experience, some games will be able to solve the security problems that exist randomly on the chain, and even regulators can intervene in a more compliant and non-interfering manner.

The imagination space is large enough, but now the TPS is really low, and the level of 2-5 TXS per second can barely approach the level of the EVM mainnet, and there is still a long way to go to reach the rollup of 10,000 transaction processing capacity.

*Note: I mainly talked about some insights from the perspective of Crypto, and the application scenarios of FHE also have great application scenarios in the field of Internet and AI, and interested friends can further study. *

Source: Golden Finance