Nvidia's KV Cache Transform Coding (KVTC) compresses LLM key-value cache by 20x without model changes, cutting GPU memory costs and time-to-first-token by up to 8x for multi-turn AI applications.

This approach can be viewed as a memory plug-in for large models, providing a fresh perspective and direction for solving the ...

MIT researchers developed Attention Matching, a KV cache compaction technique that compresses LLM memory by 50x in seconds — ...

This article outlines the design strategies currently used to address these bottlenecks, ranging from data center systolic ...

LLC, positioned between external memory and internal subsystems, stores frequently accessed data close to compute resources.

A magnetic tunnel junction engineered to produce four distinct resistance states instead of the standard two could double the data density of spintronic memory without requiring additional physical ...

VAST Data Federal's Randy Hayes said agencies looking to advance AI should replace fragmented systems with a single data ...

A technical paper titled “HMComp: Extending Near-Memory Capacity using Compression in Hybrid Memory” was published by researchers at Chalmers University of Technology and ZeroPoint Technologies.

At the Huawei Product & Solution Launch during MWC Barcelona 2026, Yuan Yuan, President of Huawei Data Storage Product Line, officially launched Huawei's AI Data Platform. The platform integrates ...

Marvell Technology, Inc. (NASDAQ: MRVL), a leader in data infrastructure semiconductor solutions, today announced Marvell® ...