NVIDIA and Pinterest detail collaboration running inference with vLLM
vllm_project · x · 2026-09-15
NVIDIA AI Infra announced a collaboration with Pinterest and the vLLM project, sharing a detailed write-up of how Pinterest uses the open-source vLLM inference framework, another notable production deployment case for the project.
More from Infra
- 500-page open-source AI Infra book released: quantitative systems design for the LLM era — udmrzn · 2026-09-15
- Oracle rides $300B OpenAI deal toward $1T market cap, but backlog quality risk looms — thedealdirector · 2026-09-15
- Modal unveils Runtime speaker lineup spanning AI infra, science and robotics — charles_irl · 2026-09-15
- Trump calls AI takeover a 'hoax' while touting data center buildout as biggest economic engine — kimmonismus · 2026-09-15
- Off-grid DGX Spark rig on 600W solar joins P2P training of an 80B model — jon_durbin · 2026-09-15
- Undervolting a RTX 5090 Local LLM Rig Cut 150W and Actually Raised TPS 3% — MasterNomie · 2026-09-15