Removing Bottlenecks

Today, Ethereum is almost completely translucent. This cannot continue to be the case if onchain is to become the next online. Confidentiality is the next major bottleneck for widespread adoption of Ethereum and FHE is the most ideal and ambitious solution.

Just as Fhenix is solving for the lack of confidentiality on Ethereum, so Celestia solves the data availability bottleneck for blockchains. Data Availability (DA) ensures that validators and users can access transaction data to verify the state of the blockchain and detect potential fraud. Celestia provides high throughput DA to address this problem with data availability sampling (DAS), which allows for DA to scale as more light nodes join the network.

The Future is Modular

Layer 2s scale Ethereum by separating the execution and consensus layers in order to facilitate the fast execution of transactions and computations. DA solutions like Celestia extend the modular approach to DA, reducing excessive fees while increasing the potential for the types of apps that developers can build. The modular approach provides a flexible solution to horizontal scaling for Ethereum, removing constraints while maintaining the security guarantees of the underlying L1.

Fhenix itself is also inherently modular by design. It separates the computationally expensive FHE execution from consensus in order to provide confidentiality while retaining the security guarantees of Ethereum.

The decision to utilize Celestia for Fhenix’s DA underscores an important strategy for our goal of bringing confidentiality to Ethereum and unlocking widespread adoption: lowering the barrier of entry for developers and making it as easy as possible for them to build confidentiality-enabled applications.

How to Apply

The Fhenix Grant program is for Fhenix projects which harness FHE to to bring confidentiality onchain or help others to do so through the creation of crucial FHE infrastructure. Fhenix grants are substantial and are intended for ambitious projects with specific, measurable milestones. Both smart contract applications incorporating confidentiality and infrastructure projects facilitating them are eligible.

To apply for a Fhenix grant, visit the Grants page at Fhenix.io and apply.

Make sure to clearly communicate how your project will benefit the Fhenix network and push the bounds of onchain confidentiality. Fhenix grant applications which stand the best chance of acceptance are thoughtful and milestone-based, with a distribution budget attached to specific development timelines.

While Fhenix grants are for larger-scale projects which utilizes FHE for onchain confidentiality, Fhenix bounties are for more specific objectives. Eligibility for Fhenix grants are open-ended, whereas bounties have smaller, specific scopes attached to them.

You can view the list of bounties, along with a description and associated amount on the Fhenix Bounty page.

Review Process

Our team will review all Fhenix grant and bounty applications on a rolling basis. If we determine more information is needed, we will reach out to schedule a call with the team to gather the additional information needed to make a decision.

We anticipate receiving many applications and will work as quickly as possible to process them. While it’s difficult to estimate the time needed to process individual grant and bounty proposals, you should receive a response in around two weeks time, at which point applicants will be notified via email whether or not their proposal has been accepted.

Where Existing Data Verification Solutions Fall Short

How do you securely and credibly prove you are who say you are digitally? It’s a good question, and one that’s only going to grow increasingly important to answer as more of our personal and financial lives move online/onchain. From proving your legal identity, to work experience, to ownership of real-world assets, the list of use cases for secure and trustless data verification is long.

Existing Web 2.0 identity verification solutions operate by placing trust in centralized third parties to securely store and process private data in order to perform data verification. The system is not only cumbersome, but vulnerable to data leaks and potential abuse from malicious actors. The solution is a decentralized system which removes the burden of verification from exploitable third parties and instead relies on math and code to execute this critical process.

Proving, Improved

The zkPass protocol guarantees private data authenticity by generating a Zero-Knowledge Proof (ZKP) in order to selectively prove various types of personal data. These zero-knowledge proof computations are performed locally, removing the need to upload sensitive data to third parties. The possible use cases for this more secure form of data verification are great, especially when combined with the confidentiality-enabling capabilities of Fhenix.

In addition to securely verifying specific pieces of data, zkPass could also be utilized to create a reputation system using the verified data points ZKPs provide as attestations for an entity’s onchain activities. Couple that with the selective confidentiality enabled by Fhenix and you have an incredibly powerful and flexible solution to onchain identity and reputation.

This collaboration opens up the design space for applications relying upon private, secure, and verified data and identity. For example, imagine a confidential DEX which uses Fhenix/FHE to obfuscate potentially personally identifiable trading history. This confidential data could then be complemented by zkPass’s data verification to prove things like trade volume and asset ownership, providing a trusted record of onchain achievement without ever introducing the possibility of exposing sensitive user data. This is just one example of the many potential use cases made possible by combining these two distinct and complementary technologies.

The Importance of FHE in the AI Age

As an ever-increasing amount of our data moves online, the need for data privacy technologies has never been greater. In this inexorable march towards digitization, a robust and trustless privacy technology is needed to safeguard our digital sovereignty. This necessity has only grown more obvious with the advent of AI and Large Language Models (LLMs), which utilize massive amounts of data for training.

Fully Homomorphic Encryption (FHE) is an ideal encryption solution for AI because it allows for direct computation on encrypted data. This enables LLMs to process the vast amounts of data needed to operate blindly, without ever needing to decrypt it. This results in a safer, more secure relationship between users and AI.

The collaboration between Fhenix and Privasea will leverage the unique talents and expertise of both parties to progress development at the intersection of AI and blockchain. These efforts include the continued development of FHE libraries, providing interoperability between the two projects, and hardware acceleration initiatives.

Software Development

Both Fhenix and Privasea will work together to extend Zama’s TFHE-rs library, a core infrastructure component of both projects. Additionally, both teams will work to enable the integration of Privasea applications on top of Fhenix’s Layer 2 infrastructure.

Fhenix and Privasea will also explore ways to integrate additional homomorphic encryption schemes based on CKKS/BGV/BFV, which allow data packing and SIMD parallel processing. This will provide better support for large-scale, high-precision computing scenarios, widening the product possibilities for both parties.

Hardware Acceleration

Alongside these software development initiatives, Fhenix and Privasea will also work in tandem on potential avenues for hardware acceleration. Specifically, we’re looking at hardware acceleration on the underlying NTT/FFT using high-parallelism, high-performance hardware such as GPUs and FPGAs. This hardware advancement is expected to provide a significant breakthrough in the efficiency of fully homomorphic encryption algorithms. We’re also taking a look at programming custom ASIC chips to improve Fhenix’s performance.

Why are we building Fhenix?

Ethereum has made huge strides over the last decade, with some of the most recent and significant advancements including the move to proof of stake and the maturation of Layer 2s as an effective scaling solution. But a glaring obstacle still remains for the world’s largest smart contract platform: it has no confidentiality.

Ethereum is well on its way to becoming a network capable of scaling human coordination in the digital age, but its current lack of data encryption presents a serious impediment to its growth trajectory. The promise of Ethereum is one of a permissionless network which returns sovereignty to the people who use and maintain it–the next, upgraded iteration of the internet. In order to realize this vision it must possess a robust and flexible answer to the question of confidentiality, a fundamental building block of human orchestration. Ethereum cannot reach its next stage of maturity without the ability to securely safeguard confidential data in a scalable manner.

We’ve spent years in the cryptographic and blockchain space researching the ideal solution to this problem and we came out the other end with a strong conviction that FHE is the answer. 

The key differentiator that sets FHE apart from other privacy enhancing solutions is the ability to perform computation on encrypted data. This simple yet powerful distinction has huge implications. It means data, which has become ever more valuable in the digital age, need not be decrypted for processing, making it that much more secure.

For a long time production-scale use of FHE wasn’t feasible due to its high computational expense, but advancements in both FHE and computer hardware have created a new reality. FHE is ready for prime time, and Fhenix is here to lower the barrier to entry of this exciting new technology by providing an accessible, EVM-compatible network for application developers to build on.

Come and build

This is a call to all hackers, cypherpunks, crypto nerds, and would-be tech entrepreneurs.

If crypto is the vanguard of cutting-edge software development, FHE is at the forefront of this technological revolution. This presents an unprecedented opportunity to build an entirely novel set of applications on Ethereum by virtue of the confidentiality that FHE enables.

Being at the forefront is not for everyone. We are in the early stages of confidential smart contracts. Many dismissed the internet, mobile computing, and Bitcoin until their network effects were already too large to ignore.

If it were easy, everyone would do it.

Fhenix is for contrarians with conviction–optimistic, heterodox builders with novel, visionary ideas who stand out in a sea of derivative, copy-paste GitHub repos. A blue ocean of applications awaits those with the foresight and conviction to see this bigger picture.

We’re here to support you with the requisite infrastructure, documentation, and dev tooling to help you turn your dreams into reality. Fhenix is fully EVM-compatible and intentionally built to be as developer-friendly as possible. Solidity-based developers can immediately begin building with familiar tools like Hardhat, Remix, and more. The goal is to make getting started with Fhenix as easy as possible.

We’ll also be offering generous grant and bounty programs to incentivize this new school of builders. The details of these programs are still being hashed out (pun intended), but you can apply for early access through the Fhenix docs.

In the meantime, fire up your development environments, because the first iteration of the Fhenix testnet is officially open and ready for deployment!

A lot of work has gone into this initial version of the Fhenix testnet and we expect to make even more improvements based on your feedback and ideas. We can’t wait to see what kind of interesting use cases you cook up in this new, confidentiality-enabled design space.

How Fhenix’s FHE Coprocessors Works

A coprocessor serves as a companion processor designed to offload specific computational tasks from a host chain – whether it’s Ethereum, an L2, or an L3, to a designated processor living outside the scope of the host.

While different types of coprocessors exist, Fhenix’s FHE coprocessors have a specific focus on maintaining data confidentiality and leveraging FHE rollups. FHE rollups have built-in encryption and unlock many novel use cases. For example, confidential AI is possible by having FHE coprocessors run intensive computations away from the host chain while ensuring that the input data is kept private.

FHE coprocessors drastically improve the efficiency of FHE-based operations, while maintaining the same level of security as the host chain.

Interacting with an FHE coprocessor is simple: an application on the host chain invokes the coprocessor, asking it to perform some specific computation over a set of encrypted inputs, and gets the result back, which it can then use onchain.

Computations are secured through Fhenix’s optimistic rollup architecture, employing fault proofs to ensure data integrity. However, these fault proofs necessitate a lengthy dispute period (normally a seven day window) for finality, which is impractical and severely limits the potential of FHE applications.

EigenLayer helps to address this challenge. It provides a ‘fast-lane’ mode in which operators verify the execution and attest to its correctness. These operators are secured by sufficient stake, and can later be slashed if they act maliciously. It’s important to note that having a dispute window is still critical, as it allows these operators to remain accountable after-the-fact, while sufficiently ensuring that no profit-maximizing operator misbehaves. In other words, thanks to EigenLayer’s cryptoeconomic security, apps utilizing an FHE coprocessor can use their output as soon as the operators verify execution, thereby facilitating substantial scaling of confidential onchain computation.

New Applications Horizon

The versatility of FHE coprocessors spans a wide range of scenarios, especially in domains prioritizing confidentiality.

For instance, consider their application in onchain auctions, where bid confidentiality ensures fairness and a more economical outcome for all bidders. Similarly, with onchain AI, FHE coprocessors can run intensive computations outside their host chain while still ensuring that the input data is kept private. The list of potential use cases is extensive, including decentralized identity, confidential DeFi, encrypted gaming, MEV protection, and many more.

To learn more about FHE use cases, click here.

A Future Shaped by FHE Coprocessors

Fhenix’s FHE coprocessors offload computational tasks and accelerate the rate in which confidential onchain transactions can be processed. Given that security is inherited from Fhenix’s optimistic rollup design, an existing limitation is that optimistic fault proofs are subject to a seven-day dispute period.

This partnership between Fhenix and Eigenlayer marks a significant milestone in the evolution of blockchain technology. By harnessing the power of FHE coprocessors, developers will be able to build applications that utilize confidential data efficiently while maintaining the same level of security. This technological advancement unlocks new capabilities for blockchain applications that were not possible before.

Challenges with Public DeFi:
Many areas of DeFi hold tremendous promise, including asset tokenization, derivatives, and payments. Tokenization alone represents a $16T opportunity, and institutions are realizing that DeFi isn’t going away.

That being said, the industry is early and there are inherent issues in having a fully transparent financial system. These include:

• Privacy concerns: Many transactions may entail sensitive data, such as institutional payments, or anything relating to personal data. This will become an event greater concern as analytical tools develop, simplifying the task of tracking one’s wallets and potentially making it easier to correlate one’s wallet to their identity.

• MEV extraction: Front-running and other forms of MEV extraction are common in DeFi, creating a hidden tax on transactions.

• Regulatory challenges: DeFi is facing increasing regulatory scrutiny, likely to only accelerate.

Fully Homomorphic Encryption (FHE) enables computations to be done on encrypted data, which was never before possible. This means that on-chain data can be encrypted while still updating the network’s shared state transaction-by-transaction.

From a privacy perspective, encryption keeps all data confidential and auditable, which coincidentally is critical for complying with data protection and privacy laws. It also means that block proposers cannot view transaction data, therefore prioritizing based on network rules rather than personal gain and preventing MEV.

Next, let’s take a look at technological advancements and considerations.

FHE in DeFi: Technical Considerations:
A challenge with FHE in DeFi is that computations must be exact, which requires the use of a specific type of scheme. A scheme refers to a standardized and systematic approach that helps developers manage data or code to efficiently solve problems.

The best DeFi-specific FHE schemes are known as BGV or BFV, which we covered in our blog post titled FHE Challenges: Part 1. While other schemes may also be used, BGV and BFV are considered the most efficient as they batch multiple computations together. These schemes are still being developed alongside FHE libraries and compilers, which in combination will radically improve the development experience.

Meanwhile, the fact that the tech is so cutting-edge means that it’s quite challenging to do computations on ciphertext (encrypted data). For example, in the past only the denominator in a division was able to be ciphertext. Recent advancements have made it possible to divide a ciphertext by another ciphertext, opening up the potential for more types of confidential DeFi transactions. AMM calculations are a use case, as encrypted calculations on DEXs (like Uniswap) require the ability to divide ciphertexts, which wasn’t formerly possible.

In the future, it’s likely that FHE tooling will improve further and open up the potential for more types of transactions, while the actual computing process also becomes significantly more efficient.

1. Confidential Transactions

Blockchain’s innate transparency imposes many limitations on potential use cases and adoption by exposing sensitive transaction details. Payments, voting, medical records, one’s financial history, and many other use cases require confidentiality, and the list will only expand alongside blockchain adoption. By enabling confidential and yet auditable transactions to take place on-chain, both consumer and corporate users of the blockchain users can benefit greatly.

Benefits include:

• Privacy protection: complete privacy over user identities, transaction amounts, and other sensitive information. This not only protects the mempool (encrypted state), but also the fields used to advance the networks state (encrypted transit).
• Enhanced security: transactions remain tamper-proof and private throughout the entire process.
• Preventing manipulation and front-running: full transaction privacy reduces the potential for unethical practices while ensuring fair transaction processing.
• Regulatory compliance: easy alignment with data protection and privacy laws, regardless of jurisdiction.

2. On-Chain Auctions

On-chain auctions, ranging from traditional bids to NFT marketplace offerings, DeFi liquidations, governance decisions, and token sales, can significantly benefit from Fhenix’s confidential smart contracts. These contracts allow users to place encrypted bids, maintaining anonymity and fairness without exposing personal data or sensitive bid details. Upon the auction’s closure, only the winning bids are disclosed, an approach especially valuable for high-stake or sensitive transactions involving premium items or significant financial decisions.

Benefits include:

• Encrypted Bids: Bids are encrypted to ensure confidentiality and prevent exposure during the auction process. 
• Full Smart Contract Functionality: An automated auction process with full confidentiality, including bid submission, validation, and winner selection.
• Reveal Mechanism: Secure reveal winning bids at the auction’s conclusion without compromising the integrity of the process.
• Privacy-Preserving Protocols: Maintain confidentiality of participants’ identities, bid amounts, and other confidential information.

3. Confidential Voting

On-chain voting is critical for Decentralized Autonomous Organization (DAO) voting, as well as for consensus mechanisms and any other forms of on-chain voting where confidentiality and security are needed.

Projects using Fhenix’s FHE can vote with complete confidentiality, while the system knows that their votes are entirely valid. FHE’s stringent security & privacy characteristics also make it regulatory compliant, enabling voting even in high-stakes environments, such as governmental elections.

Benefits include:

• Voter Privacy: This protects voters from external influence or coercion, and safeguards their privacy.
• Security: Encryption makes it challenging for attackers to trick the system, study voter patterns, or manipulate the system in any manner.
• Regulatory Compliance: Adherence to election laws and privacy regulations, especially critical for important vote situations.

4. Confidential Gaming

Multiplayer Web3 games face various challenges caused by the blockchain’s transparency, including exposing real-time positions, analyzing historical strategies, in-person identity correlation, and other forms of potential manipulation. For example, on-chain poker is challenging as all cards are publicly visible and the game can be manipulated.

With FHE, developers can integrate confidentiality into Web3 games, keeping all data entirely confidential while still updating the blockchain’s state after each move.

Benefits include:

• Securing In-Game Transactions: Public blockchain records enable opponents to analyze past strategies and pending moves or to manipulate the game such as through front-running actions. Encrypting both financial transactions and player actions protects against real-time manipulation.
• Protecting Player Data: Player data, such as financial assets and personal information, can be discoverable and expose them to phishing attacks and real-world security risks. Encryption anonymizes player activity, reducing these risks.
• Transparent State Updates: Existing solutions such as ZKPs have transactions processed off-chain and only periodically updated on-chain. FHE has all in-game actions encrypted on-chain, yet verifiable via decryption.

5. Decentralized Identity (DiD)

Decentralized Identity (DID) represents an on-chain identifier facilitating a secure and verifiable digital identity. This approach ensures that an individual’s identity remains self-owned and isn’t subject to control by a central authority.

DiD at its core holds sensitive personal data that will be targeted by malicious parties. On-chain FHE enables the secure storage and verification of encrypted user data, while also providing selective disclosure and control over key parameters.

Benefits include:

• Control Over Personal Data: Selective disclosure over identity attributes, for example, proving citizenship while ensuring all other details remain fully secure.
• Secure Authentication: Traditional identity systems are frequently subject to data exploits. Fhenix protects login credentials and authentication tokens from being intercepted or duplicated.
• Interoperability and Data Portability: DIDs operate across various platforms and services. Encryption ensures that while this data is portable and interoperable, it remains confidential and can only be accessed by entities with the appropriate decryption capabilities.

6. MEV Protection

Pending blockchain transactions are publicly visible in the network’s mempool, which can lead to value extracted through methods such as front-running or sandwich attacks. For example, a pending profitable transaction could be front-run by another party submitting the same transaction with a slightly higher gas fee paid. This is known as Maximal Extractable Value (MEV) and it’s a hidden tax.

This is easily preventable through Fhenix’s built-in confidentiality, whereby pending transaction data is kept confidential.

Benefits include:

• Preventing Front-Running and Exploitation: By having the public data of all transactions encrypted and hidden from those monitoring the mempool, MEV bots to cannot engage in front-running, sandwich attacks, or other predatory practices.
• Ensuring Transactional Fairness: Block proposers cannot view transaction data, therefore prioritizing based on network rules rather than personal gain.
• Protecting Against Censorship: An inability for block builders to censor transactions.

7. Privacy-Preserving A.I.

The application of A.I. is expanding rapidly across various sectors, utilizing extensive amounts of data to perform intricate computations.  FHE allows AI models to process encrypted data, enabling the use of sensitive personal or financial information in AI computations securely, without privacy or security risks.

For example, Ethereum Founder Vitalik Buterin’s recent blog post covered how the overhead cost of putting AI inside cryptographic boxes is quite low, and FHE is a viable solution to this.

Additionally, Zama recently updated their Concrete-ML library to add the ability to train data using an encrypted ata set. They also plan to make have their Concrete-ML library be compatible with TFHE-rs ciphertexts, which is their Rust implementation of the TFHE (Zama’s variation of FHE).

The result is that quick advancements are being made when it comes to privacy-preserving AI on-chain, and that languages and libraries will be increasingly easy for developers to use.

The main implications are:

• Data Privacy: A.I. models can use fully encrypted training data, while ensuring that it remains encrypted until it reaches its intended destination.
• Security: Full encryption prevents unauthorized access or data manipulation.
• Regulatory Compliance: Facilitates adherence to stringent data protection regulations in AI applications.

On-Chain Auctions

On-chain auctions use smart contracts to manage and execute an auction, from the initial asset listing to determining the winner and finalizing the transaction.

Auctions are prevalent in the crypto sphere due to factors like the market’s high volatility, which frequently results in the liquidation of positions, the continuous, round-the-clock trading of diverse assets, and the relatively low barriers to entry for participants.

Each category of on-chain auction alone represents a vast sum of funds.

For example, December 11th, 2023 alone saw over $320M in positions liquidations occur, and while a significant fraction took place on centralized exchanges, trading is increasingly moving on-chain. Meanwhile, ICOs, tending to use auctions, have raised over $50B since 2017.

On-chain carbon credits, NFT marketplace sales, and many more areas of blockchain will require well-designed blockchain infrastructure with privacy & security built-in.

Issues with On-Chain Auctions

On-chain auctions currently struggle with 3 primary challenges:

1. Privacy
2. Integrity
3. Compliance

Privacy is an issue as on-chain auctions can be correlated to one’s real-life identity, making bidders targets for scams. For instance, Mark Cuban recently lost nearly $900,000 to a phishing attack, with many other noteworthy phishing attacks such as one against a key Harmony employee that led to the loss of $100M in funds. Additionally, dissecting one’s on-chain actions can lead to strategic bids made based on one’s identity or other factors. We also believe that blind auctions lead to more efficient price discovery, benefitting the entire market.

Integrity is an issue as publicly visible bid details on the blockchain could lead to issues such as front-running, collusion, or other forms of manipulation. For instance, existing or impending bids could be monitored, with a competing being placed at the very last second for a cent more. Encryption keeps bids secret until the auction concludes, creating a level playing field for all participants.

Last of all is regulatory compliance. Certain jurisdictions have strict privacy and data protection laws that require data confidentiality, and strict compliance around regulated assets. This will also increase as more institutions arrive on-chain, as well as regulated assets such as tokenized commodities, real estate, and financial instruments. For example, auctions revolving around regulated assets necessitate strict privacy and security measures, possible through Fully Homomorphic Encryption.

FHE on the Blockchain

Addressing the above challenges and ensuring the complete encryption of on-chain auctions is extremely easy to implement using Fhenix’s FHE Rollups.

FHE Rollups enable developers to craft custom application chains with built-in FHE capabilities while using well-known EVM languages such as Solidity. For example, an NFT marketplace could be its own FHE rollup that easily connects to both external EVM and non-EVM networks, while having all transactions be fully confidential. Concurrently, another interoperable rollup could be centered around fully encrypted Initial Coin Offerings (ICOs).

Developers can easily deploy any custom-purpose Layer 2 with privacy, security, and compliance built-in from the core while knowing that transactions are auditable if need be.

What about drawbacks?

One challenge is that FHE is cutting-edge, meaning that it’s also quite new.

In our 2-part series, we covered 5 separate challenges:

• Part 1 discussed nascent FHE schemes, libraries & compilers, and a need for advanced threshold decryption.

• Part 2 discussed the need for ZKPs for both user inputs & computations, data availability requirements, and the need for improvements in FHE hardware & software.

These technical aspects still require some modification, but progress has been moving extremely quickly and we expect the same for 2024.

Challenge: Implementing Zero-Knowledge Proofs for Verifiability

While FHE enables arbitrary computation over encrypted data, it cannot prove that something is correct without revealing the underlying information. This is where zero-knowledge proofs (ZKPs) excel, and combining both is useful in three ways:

1. Proof of Plaintext Structure: ZKPs can be used to demonstrate that the input plaintext is properly structured, which impacts the ciphertext (encrypted data) conforming to the encryption scheme. This cannot be done by the typical SNARK/STARK approaches common for ZK rollups because FHE’s lattice-based cryptography is post-quantum secure, and instead requires a proof system designed for lattice-based relations.

2. Proof of Plaintext Input: ZKPs can also prove that the plaintext is not only properly structured, but that it meets the application’s requirements. For example, I cannot send more funds than I have in my wallet. zk-SNARKs can be used here, but the challenging part is confirming that the input for this system is the same as the input for the plaintext structure system.

3. Proof of FHE Computation: ZKPs can be used to prove that computations are correct.

Solutions:
Each of the above 3 is in the midst of ongoing research providing promising results:

a. ZKPs of plaintext structure require a cost-efficient proof system for lattice-based relations, which is an area of ongoing research.

b. ZKPs of plaintext input are being advanced by firms such as Mind Network which have been integrating ZKPs to ensure zero-trust inputs, alongside the use of FHE. Sunscreen’s ZKP Compiler also shows promise here through the use of Bulletproofs, although they are not the most efficient proof system.

c. ZKPs of correct computation can be addressed via “Verifiable FHE” whereby a validator submits a ZKP to prove honest transaction execution. Assuming others can verify the ZKP was done correctly, only a single part is needed for execution. SherLOCKED is an example of an early prototype in this space, which offloads FHE computation to Risc0’s Bonsai zkVM and returns a ZKP.

Each area is advancing rapidly and we will be providing updates via our Twitter on the latest information.

Challenge: Low-Performance FHE Hardware

Latency is a recurring challenge for FHE and this is no different when it comes to hardware.  FHE computations over encrypted data were 100,000x slower than over plaintext data, but newer encryption schemes and FHE hardware advancements mean that FHE computations are now only 100x slower.

The fundamental reason for this is that calculations involve complex polynomial computations and time-consuming ciphertext operations. It’s also harder to accelerate blockchain-based FHE compared to application-specific use cases (such as Machine Learning FHE) as it requires the ability to process more general computations versus a more niche set.

Solutions:
Improvements are being made on 3 fronts:

a. FHE Hardware: FHE Accelerators are specialized hardware devices that enhance the performance of FHE computations by speeding up the complex mathematical operations involved. One example is the Fixed Point (FPT) accelerator that can exploit FHE’s inherent noise through the implementation of TFHE bootstrapping. Another example is BASALISC, an architecture family of hardware accelerators that use the cloud to accelerate FHE computations and is the first to implement the BGV scheme.

b. Optimized Software: F1 and CraterLake are two FHE Accelerators that combine both hardware and software elements to increase performance, and others may follow suit in the future. Fully functional compilers will likely be developed, with this software capable of optimizing FHE programs dependent on the specific For HE scheme chosen.

c. Scaling Out: Currently, FHE hardware accelerators are focused on improving individual accelerators vertically, rather than connecting various ones horizontally. This could drastically improve performance and is comparable to proof generation with ZKPs, where multiple proofs can be generated in parallel. A challenge with this is data inflation: data size increases drastically during encryption and computation, making it harder to connect various FHE hardware accelerators. Therefore, a promising area of research will be to look into the networking infrastructure amongst FHE hardware accelerators.