Exploratory AI infrastructure project modeling semantic KV-cache orchestration, memory tiering, and HBM/CXL movement tradeoffs for long-context LLM inference.
compression sram dram dma hbm memory-hierarchy cache-management prefetching systems-architecture kv-cache runtime-systems cxl long-context llm-inference ai-infrastructure systems-research inference-systems memory-orchestration serving-systems accelerator-research
-
Updated
May 19, 2026 - Python