Case Study

Diagnosing Inference Server Failures

An inference server is processing multiple text generation requests concurrently. The system monitor shows that 40% of the total memory is free. However, the server frequently fails to start processing new requests that require long contexts, reporting 'out-of-memory' errors, while shorter requests are still processed successfully. The system's memory manager allocates a single, uninterrupted block of memory to store the cached information for each individual request. Based on this allocation method, what is the most likely cause of this discrepancy between available memory and allocation failures?

0

1

Updated 2025-09-26

Contributors are:

Who are from:

Tags

Ch.5 Inference - Foundations of Large Language Models

Foundations of Large Language Models

Foundations of Large Language Models Course

Computing Sciences

Analysis in Bloom's Taxonomy

Cognitive Psychology

Psychology

Social Science

Empirical Science

Science