pragma.vision Technology observatory & strategic foresight

Verification register Compute & Web Infra

Living definition

What is llm-d?

MLOps & AI Infrastructure Compute & Web Infra

As of
2026-08-20
Revision
2026-08-17.1
Method
v1.3.0

Definition

The term, in context

AI-assisted draft · approved dataset

llm-d is a Kubernetes-native distributed inference framework, built on the vLLM engine, that splits large-language-model serving into separately scaled pools for the compute-heavy prompt-processing phase and the memory-bandwidth-bound token-generation phase. It routes each request toward the replica most likely to hold its relevant cache.

Live readiness status

Status as of the current dateline

AI-assisted assembly · derived results

As of 2026-08-20, verified readiness is 52 (claimed 70, reported 61, gap 18) — Critical evidence strength; current signals suggest Track; not yet.

The readiness fact belongs to the canonical Readiness Verdict for llm-d.

Dataset approval

Human editorial release

pragma.vision editorial — standing authorization (operator dev@soft.house, 2026-07-17)

Dataset
2026-08-17.1
Hash
sha256:0c37108a182f510fc8cf4d1dfc3a30339f2b59e62654201edea2f530a268707e
Approved
2026-08-17

Your opinion

Tell us anything.

What works, what doesn't, what's missing — especially about our watches, lenses, and the register itself. Anonymous is fine; leave an email if you'd like a reply.