Open App →
Back to News Feed
Digitimes September 15, 2026 By DIGITIMES neutral

Astera Labs targets the KV cache bottleneck as agentic AI outgrows GPU memory

AI / LLMNVIDIA / GPUMemory
Astera Labs has added three products to its Leo memory controller line, aimed at a problem that arises as AI inference shifts from single queries to agents that run continuous loops: the key-value cache outgrows the HBM on the accelerator, and the usual fallback makes things worse.
Read original article ↗

Related Articles

Delos Data Is Reimagining the Interconnect for the AI Age

Multiple AI models. Heterogenous compute. A variety of network topologies. Copper and optical. The mix of silicon, stora

HPC Wire · September 15, 2026

Scaling Federated Learning Across Docker, Kubernetes, and Slurm with NVIDIA FLARE

Federated learning (FL) projects often begin with a straightforward setup: one server, a few clients, and one dataset at

Nvidia Developer Blog · September 15, 2026

Meta’s new One subscriptions put a price on social media and AI

Shortly after launching its new do-everything AI assistant Muse, Meta's launching subscription bundles that pair its sta

The Verge · September 15, 2026

Delos Data Closes over $100 Million to Deliver Delos Nonstop AI™ for Faster, Stronger, More Efficient AI Capacity

PALO ALTO, Calif., Sept. 15, 2026 /PRNewswire/ -- Delos Data today announced it has raised over $100 million from Matrix

PR Newswire · September 15, 2026