According to Foresight News, ZhenFund announced a donation to the open source AI project vLLM. vLLM was developed by a three-person team at the University of California, Berkeley, and aims to build the fastest and easiest-to-use open source LLM inference service engine. Other donors to vLLM include a16z, AMD, Anyscale, AWS, Crusoe Cloud, Databricks, Deepinfra, Dropbox, Lambda Lab, NVIDIA, Replicate, Roblox, RunPod, Sequoia Capital, Trainy, UC Berkeley, and San Diego.