vLLM Hit 7.8x More Per-GPU Throughput on Nvidia’s Rubin
vLLM already runs on Vera Rubin NVL72 at 7.8x the per-GPU throughput of GB200, and memory bandwidth explains most of the jump.
The Cheapest Inference on the Market Is a Classifier, Not a Chat Model
Microsoft-Decision-1 lands in a crowded decision model market. The pitch is not accuracy, it is $0.042 per million input tokens with output free.
Every AI Agent Should Get Its Own Kernel. Edera Built That on Kubernetes.
Edera shipped Project Lunchbox, three primitives that give AI agent isolation on Kubernetes real teeth. Every agent gets a microVM and a fork.
Broadcom Wants VMware to Be Where Your AI Runs. Inside Your Own Walls.
Broadcom's VMware AI Factory turns VCF into a private cloud for production AI. It is a direct play for the token economy.
K8sGPT Turns Cluster Error Messages Into Fixes a Human Can Use
K8sGPT applies AI to Kubernetes troubleshooting, explaining cluster failures in plain language and suggesting a fix.
Spacelift Enlists an AI Copilot to Review Terraform Before You Do
The Spacelift AI copilot reviews Terraform and OpenTofu changes before a human looks, catching risky plans and suggesting likely fixes.