Benchmarking Post-Quantum Key Encapsulation in S3-Compatible Object Storage under Client Concurrency
ML-KEM object encryption cuts S3-store throughput by about 4-5% only when client CPUs are saturated.
The paper benchmarks per-object ML-KEM key encapsulation against classical key exchange on an S3-compatible object store, using two client implementations, two processor types and up to 500 concurrent workers. Throughput fell only when the client processor was at or near full load, by 4.0 to 4.9 percent in the main client, with one unexplained exception. Controlled tests traced most of that loss to how the software library was called, and a second native client showed the size of the loss depends on the implementation. The lasting cost is storage: 1,056 extra bytes per object versus classical key exchange.
- Tests used two clients, two CPU types and up to 500 workers.
- Main-client throughput dropped 4.0-4.9% only near full CPU load.
- Much of the slowdown came from library call overhead, not ML-KEM itself.
- Each object stores 1,056 more bytes than with classical key exchange.
From the original article
1 min read at arxiv.orgData encrypted today with classical public-key algorithms could be decrypted once large quantum computers exist. Cloud object stores keep data for years, so the client-side encryption that protects it must move to post-quantum algorithms such as ML-KEM, which adds a key-encapsulation step and a larger ciphertext to every stored object.
By Revanth Reddy Lingala · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.