FHE vs Secure Enclaves: Comparing Privacy Technologies

FHE and secure enclaves (like Intel SGX and ARM TrustZone) both enable computation on sensitive data. However, they take fundamentally different approaches with different security properties. Understanding these differences helps you choose the right technology.

Fundamental Approaches

FHE (Fully Homomorphic Encryption)

Mathematical protection: data remains encrypted during computation, security based on cryptographic hardness.

Secure Enclaves

Hardware protection: data decrypted inside a protected hardware region, security based on physical isolation.

Security Model Comparison

FHE Security

• Based on mathematical problems (lattice hardness)
• No trusted hardware required
• Post-quantum secure (lattice-based schemes)
• Data protected even if server is fully compromised

Enclave Security

• Based on hardware isolation
• Requires trust in hardware manufacturer
• Vulnerable to side-channel attacks (Spectre, Meltdown variants)
• Attestation proves code integrity

Attack Surface

FHE Attack Vectors

• Cryptographic breaks (theoretical, none known for standard schemes)
• Implementation bugs in FHE libraries
• Key management failures

Enclave Attack Vectors

• Side-channel attacks (many demonstrated)
• Hardware vulnerabilities
• Microcode/firmware attacks
• Supply chain attacks on hardware

Performance Comparison

FHE Performance

• Significant overhead compared to plaintext
• Improving rapidly (10-1000x improvement in recent years)
• Simple operations: microseconds to milliseconds
• Complex operations: seconds to minutes

Enclave Performance

• Near-native speed for computation
• Overhead mainly in enclave entry/exit
• Limited memory (historically)
• Context switches are expensive

Practical Considerations

Choose FHE when:

• You can't trust the computing infrastructure
• Post-quantum security is required
• Data must never be decrypted outside your control
• Computation is relatively simple

Choose Enclaves when:

• Performance is critical
• Complex arbitrary code must run
• You can trust hardware attestation
• Side-channel mitigations are acceptable

Hybrid Approaches

The best solutions often combine both:

• FHE for long-term data protection
• Enclaves for performance-critical operations
• FHE key operations inside enclaves
• Defense in depth

H33's Approach

We primarily use FHE for biometric matching because:

• Biometric data has lifetime sensitivity
• Users shouldn't trust any third party with their biometrics
• Our optimizations achieve acceptable performance
• Post-quantum security future-proofs the protection

Both FHE and secure enclaves have their place. Understand their trade-offs to make the right choice for your application.