SDAA498 September   2026 AM62L

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1Introduction
    1. 1.1 Platform
  5. 2Yocto Integration (TI SDK Wrynose)
    1. 2.1 Layer Stack
    2. 2.2 OpenSSL 3.5 Native PQC: No Extra Steps Required
    3. 2.3 liboqs / oqs-provider Integration
    4. 2.4 Recovering From Build Errors
    5. 2.5 wolfSSL Integration
      1. 2.5.1 Choosing a Path
      2. 2.5.2 Step A1 — Option A: meta-wolfssl
      3. 2.5.3 Step A2 — Option B: meta-oe wolfssl (Library Only)
    6. 2.6 Step B — Build
    7. 2.7 Step C — OpenSSL Provider Configuration
  6. 3Performance Benchmarks
    1. 3.1 ML-KEM — All implementations
    2. 3.2 ML-DSA — All implementations
    3. 3.3 FrodoKEM (liboqs only)
    4. 3.4 Hybrid KEM — TLS overhead
    5. 3.5 armasm Impact on wolfSSL Performance
  7. 4NEON and HW Acceleration Analysis
    1. 4.1 Acceleration Status per Implementation
    2. 4.2 Recommendation by Algorithm
  8. 5Algorithm Availability Summary
  9. 6References

Acceleration Status per Implementation

Implementation ML-KEM NTT ML-DSA NTT SHAKE/Keccak AES SHA-2
OpenSSL 3.5 default Pure C Pure C NEON keccak1600 asm HW HW
liboqs 0.15.0 (Yocto) aarch64 NEON NTT ✓ Pure C (no aarch64 path) NEON Keccak (XKCP) ✓ HW ✓ HW ✓
wolfSSL 5.9.1 (no armasm) Optimized C Optimized C Pure C Pure C Pure C
wolfSSL 5.9.1 (+armasm) Optimized C Optimized C NEON Keccak asm ✓ HW asm ✓ HW asm ✓

Key observations:

  • wolfSSL --enable-armasm enables NEON Keccak (sha3_asm.S) in addition to hardware AES/SHA-2 — this is the primary driver of the 60% to 100% ML-KEM speedup
  • liboqs has a dedicated aarch64 NEON NTT for ML-KEM — but wolfSSL's NEON Keccak advantage on SHAKE outweighs it, giving wolfSSL+armasm the overall lead on ML-KEM
  • liboqs has no aarch64 NTT for ML-DSA (AVX2 only in liboqs 0.15.0) — wolfSSL optimized C wins on ML-DSA regardless of armasm
  • Neither liboqs nor wolfSSL can use SHA-3 hardware (Cortex-A53 lacks it) — NEON Keccak is the best available for SHAKE