Accelerated Quantum Supercomputing and Post-Quantum Cryptography
summary
A GPU-based PQC library implements ML Cam and mlsa (distributed as an object file, not FIPS-tested or fully open source) and targets high-throughput TLS workloads by batching operations on GPUs. Benchmarks report key generation up to 10–20M/sec, encapsulation 8–20M/sec and decapsulation 7.5–18M/sec, with batching yielding roughly 150–320× speedups versus a single CPU core; signature batching hits millions/sec for gen/verify ranges. GPUs introduce noise that reduces some side-channel risks, and PCIe transfer and register pressure are the main engineering limits.
AI-written summary · about
chapters (13)
- 0:00Welcome
- 0:42Accelerated quantum supercomputing
- 2:09Quantum computing revolutions
- 4:58AI in quantum acceleration
- 7:56AI in quantum hardware
- 10:27AI-driven quantum optimization
- 12:14Post-quantum cryptography urgency
- 13:58NIST transition timeline
- 15:23GPU for high-throughput crypto
- 17:39PQC algorithms on GPUs
- 19:09PQC performance on GPUs
- 22:20PQC vs CPU performance
- 22:44Q&A
related talks
21mCUDA-QX and Artificial Intelligence for Quantum Computing
45mQuantum application development for multi-processor systems using CUDA-Q
21mHybrid Quantum Computing for AI
01h 07mQuantum architecture for utility scale with HPC, AI and Quantum
16mNext Level Quantum
11mScience & Industrial Innovation With Quantum-Accelerated Highperformance Computers
19mQuant-LLM: Accelerating the Serving of Large Language Models via FP6-Centric...
56mFireside Chat: The coming interplay of High-Performance Computing & QC