Building more efficient AI with vLLM ft. Nick Hill

Red Hat CTO Chris Wright interviews Nick Hill, Senior Principal Software Engineer and a core contributor to the vLLM open-source inference project. They cover PagedAttention and KV-cache management, speculative decoding, GPU utilization, and Nick's path from IBM Watson to generative AI inference. Also published as video at https://www.youtube.com/watch?v=sf1zUidJYRc and on Apple Podcasts at https://podcasts.apple.com/us/podcast/building-more-efficient-ai-with-vllm-ft-nick-hill/id1817682331?i=1000715377281