二〇二六年九月十六日
收録 16 篇
- Reimagining advertising with AI
- omnigent-ai/omnigent
- AI for Games in the Foundation Model Era
- vLLM x Novita AI: Chord, Faster INT4 MoE for Kimi K2.x. Up to 1.3x on H200, 2.15x on Untuned B300
- Product updates: Sandbox Sidecars, new models, a refreshed dashboard, and more
- The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement
- Continual Learning Mechanisms Compose for Long-Horizon Memorization
- Atria-Dawn-Preview-Ascend-w8a8
- How we trained the fastest DSpark for Kimi-K3 using GB300 NVL72
- Gemini Live audio
- Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking
- Kimi K3 Performance Optimizations in vLLM: The Road to 2.8× Throughput
- How NVIDIA NVLink 6 Delivers Multi-Layer Resiliency for AI Factories
- Dense vs. MoE Models: Active Parameters, Throughput, and When to Choose Each
- How NVIDIA Groq 3 LPX Deterministic Execution Drives Power-Efficient High-Interactivity Inference on NVIDIA Vera Rubin
- Your Agent Aced the Task. Will It Do It Again?