Skip to content
NewA calmer, faster ZeroHour

TUTTI bootstraps CKKS in under 3 ms on CPU

Original title: TUTTI: Parallel CKKS Bootstrapping in under 3 ms on CPU

arXiv cs.CRby Robin Köstler ()1 min read
AI summary · x-ai/grok-4.7Summaries by Vani AI

On 192 threads it restores 1 to 9 levels for 4096 messages in 46 to 83.4 ms, up to 3 times faster than SHIP.

A paper introduces TUTTI, a CKKS bootstrapping procedure that improves multi-threaded latency over prior low-latency methods and, for one message, undercuts TFHE. It builds on SPRU with more parallelism, more messages, and a refined analysis of LMSS block-binary secret keys. Reported times include 4.6 ms for one message on 32 threads and 2.4 ms on an Apple M5 Pro with 14 threads.

Key facts
  • TUTTI bootstraps 4096 real messages in 46 to 83.4 ms on 192 threads.
  • A 256-message run takes 15 to 34 ms on 32 threads.
  • One message takes 4.6 ms on 32 threads, versus 7.3 ms for TFHE-rs 4-bit bootstrapping.
  • On an Apple M5 Pro, TUTTI takes 2.4 ms on 14 threads against 4.5 ms for TFHE-rs.
  • The authors say functionality, key sizes, and security assumptions still differ from TFHE.
Analyze this story with Vani AIImpact, who is affected and what defenders should do, in a new chat

From the original article

1 min read at arxiv.org

This paper introduces TUTTI, a new CKKS bootstrapping procedure that outperforms previous low-latency solutions in multi-threaded latency for any number of messages and, most notably, has a lower latency than TFHE when bootstrapping a single one.

arXiv cs.CR

By Robin Köstler · arxiv.org

Read the full article at arXiv cs.CR

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.