Inference Infrastructure & ServingMicrosoft (Patel, Choukse et al.)Microsoft

Splitwise

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

Splitwise — phase splitting

Why it matters

The first proposal in the prefill-decode disaggregation line

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/2311.18677

Added in Week 35, 2026 · See that week

On the map — ⑧ Inference Infrastructure line

  1. Previous stopTensorRT-LLM
  2. This recordSplitwise
  3. Next stopSarathi-Serve
See it on the map

Same week

MambaPaperASE describes 5 categories of manufacturing AI on its tech blogOfficialGNoMEPaperDeepSeek releases DeepSeek LLM 7B and 67BOfficial

Nearby on the same axis

SGLang paperPaperNVIDIA announces H200OfficialNVIDIA releases TensorRT-LLMOfficialLMSYS releases SGLang + RadixAttentionOfficial