MiniPIC: Efficient Position-Independent Caching System for Large Language Model Inference
Researchers have proposed MiniPIC, a position-independent caching system for large language model inference that requires fewer than 100 lines of core engine changes to vLLM. Standard prefix caching fails to reuse key-value entries when repeated inputs appear at different positions in a prompt, a common scenario in retrieval-augmented and agentic workloads. MiniPIC addresses this limitation while achieving up to 49% higher prefill throughput and reducing cached-span time-to-first-token by up to two orders of magnitude.
MiniPIC (Minimalistic Position-Independent Caching) is a new design for the vLLM inference engine that enables reuse of key-value (KV) cache entries for recurring structured inputs—such as documents or code files—regardless of where they appear in a prompt. Existing prefix caching requires identical prefixes across requests to reuse KV state, while prior position-independent caching approaches either demand large server-side code changes or store KV state off-device, introducing costly host-to-device transfer overhead. MiniPIC's approach centers on two ingredients: storing unrotated K vectors in the KV cache and applying rotary positional encodings (RoPE) dynamically per request, combined with three user-facing primitives—block-aligned padding, span separator, and prompt depend—that control hashing and causal attention structure. These changes amount to fewer than 100 lines of core engine modifications plus a custom attention backend, yet are sufficient to implement multiple established PIC methods (Block-Attention, EPIC, and Prompt Cache) within a single running vLLM instance. Evaluated on the 2WikiMultihopQA benchmark with interleaved scheduling, MiniPIC improves prefill throughput by 49% over baseline vLLM, reduces time-to-first-token for cached spans by up to two orders of magnitude, and incurs only 5.7% worst-case overhead on uncached workloads. The system also integrates natively with CPU KV cache offload implementations, broadening its practical applicability.
What's missing
The paper evaluates primarily on 2WikiMultihopQA; generalization to other benchmarks, model architectures beyond those tested, and real-world production traffic distributions remains undemonstrated. The work is a preprint and has not yet undergone peer review.
What different sources said
- arXiv cs.AICenter
MiniPIC: Flexible Position-Independent Caching in <100LOC
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.