Tech Meridian ← LIVE FEED
PROMY MERIDIAN RU

RESEARCH · RESEARCH · #3207

BlockLock (arXiv:2611.06977v1): reproducible LLM inference via fixed batch geometry

The paper introduces BlockLock, a method that preserves bitwise reproducibility for a protected request by fixing the batch geometry the request sees. It describes two modes (BlockLock-Rows and BlockLock-Pack), reports preservation of target sequences and Top20-logit bits across five dense models and three single-GPU setups, and finds BlockLock-Rows cuts latency by 53.5–55.0% and raises throughput by 115.2–122.2% versus vLLM Batch-Invariant while avoiding custom kernels and retaining shared prefill compute under a stated per-deployment contract.

KEY POINTS

  1. The paper introduces BlockLock, a method that preserves bitwise reproducibility for a protected request by fixing the batch geometry the request sees.
  2. It describes two modes (BlockLock-Rows and BlockLock-Pack), reports preservation of target sequences and Top20-logit bits across five dense models and three single-GPU setups, and finds BlockLock-Rows cuts latency by 53.5–55.0% and raises throughput by 115.2–122.2% versus vLLM Batch-Invariant while avoiding custom kernels and retaining shared prefill compute under a stated per-deployment contract.
  3. BlockLock offers a practical way to get bitwise reproducibility for individual requests with much lower latency and higher throughput than batch-invariant approaches and without custom kernel development.

WHY IT MATTERS

BlockLock offers a practical way to get bitwise reproducibility for individual requests with much lower latency and higher throughput than batch-invariant approaches and without custom kernel development.

SOURCES & TIMELINE

1