How KV Cache Changes AI Performance: Solidigm Explains the Hidden Path of Every Prompt - Tech Talks
AI Field Day 8
•
6m 48s
What really happens after you hit enter on an AI prompt? At AI Field Day 8, Stephen Foskett sits down with Solidigm’s Scott Shadley to break down the journey of AI data from user prompt to hardware infrastructure. Learn how tokens move through memory, storage, and KV cache, why GPUs spend so much time waiting, and how modern SSDs are helping reduce costs while improving AI performance. This discussion explores AI infrastructure, memory optimization, storage as a memory layer, GPU utilization, and the real-world impact of architectural decisions on large language models, AI agents, and enterprise AI deployments.
Up Next in AI Field Day 8
-
The Human Imperative, Tokenomics, and...
Stephen Foskett, organizer of Tech Field Day, moderated a delegate roundtable discussion at AI Field Day 8, providing a platform for participants to share their perspectives on the current state of artificial intelligence following the event's presentations. A prominent theme across the discussio...
-
The Next Wave of Al - Inference Outsi...
Guy Currier of The Futurum Group presented on "The Next Wave of AI," focusing on the increasing importance of inference outside hyperscale data centers. He highlighted a maturing AI market characterized by a significant shift in investment from AI training towards AI inference. This trend, suppor...
-
Standardizing Gen Al Service Evaluati...
David Kanter detailed the ongoing evolution of MLPerf benchmarks, which have been an industry standard for seven years. He highlighted the need for fundamental changes, particularly in the visualization of results, moving from an outdated, spreadsheet-like format to a more modern and understandab...
