🤖 NVIDIA ships first agent CPU Vera, Huang says AI demand on parabolic climb
Audio in Mandarin Chinese · English transcript below
⚡ NVIDIA’s Vera CPU just launched, AI demand shooting parabolic, RAG crushed by long context, DeepSeek V4 tops the charts, open-source agent bootcamp, college kids fear job loss, addict-AI sued for big bucks, Gemini sneaks in an upgrade…
The explosive growth of Agent AI is comprehensively reshaping AI infrastructure to address what Jensen Huang terms "parabolic" demand. NVIDIA has delivered its first Vera CPU, specifically designed for Agents, to meet their unique compute workloads. Notably, context architecture is replacing RAG, and Redis has launched its Iris platform, together providing innovative solutions for the large-scale deployment of Agent AI.
Today's Top 3 Headlines
- AI Industry News
🤖 NVIDIA ships 88-core Vera CPU: 1.2 TB/s for Agent AI
NVIDIA VP Ian Buck delivers first 88-core Vera CPUs to Anthropic, OpenAI, SpaceXAI, Oracle Cloud: +50% per-core speed, 1.2 TB/s memory. Agent builders gain CPU-heavy sandboxes & tool calls.
Source ↗ - Technology
🤖 Jensen: AI demand soars parabolically; 5k firms onboard Dell AI Factory
NVIDIA CEO Jensen Huang told Dell Tech World AI demand is “fully parabolic,” unveiling Vera Rubin NVL72; 5,000 firms already run AI workloads on Dell AI Factory. For compute suppliers it means orders can…
Source ↗ - AI Industry News
🤖 Redis debuts Iris: context-first RAG killer, flash cost -90%
VentureBeat: Redis unveils Iris for AI Agents, swapping RAG for context fabric, flash cost 1/10 RAM, live Oracle CDC sync. Agents now pull data on demand and remember across sessions, cutting infra spend 90%.
Source ↗
+7 more headlines
- 🤖 DeepMind’s Hassabis quietly held early Anthropic stake
- 🤖 Addictive AI emerges as new corporate risk
- 🤖 1/3 UK students: AI job loss will spark social unrest
- 🤖 DeepSeek V4: 40% faster, 30% more accurate language tasks
- 🤖 OpenAI API: Build a Planning, Tool-Calling, Memory, Self-Critique Agent
- 🤖 Gemini’s stealth drop: YouTube → instant AI summary player
- 🤖 SGD drift slows LLM training; Adam speeds it 1.5×