Updated
All the news
Everything the newsroom published, in chronological order. Each item carries origin, sources and reading time.
12Today
8Yesterday
48This week
Mon, Aug 10
12 stories
CHIPS TSMC CoWoS 5.5x reticle yields exceed 98% as AI packaging capacity grows 80% this year 
BREAKING DeepMind's Hassabis steps down as CEO, Gemini co-leads Jeff Dean and Oriol Vinyals exit to start Discovery Loop 
MARKET AI data center demand drains global memory supply; gaming GPU prices double amid GDDR7 shortage 
BREAKING Google demotes Demis Hassabis as DeepMind CEO; Jeff Dean departs Google to start AI startup 
CHIPS Samsung hits 80% HBM4 yield, plans to triple Q3 revenue and challenge SK Hynix dominance 
RESEARCH CreativeInstruct Recovers Diversity Without Slowing Inference 
BREAKING Moonshot Kimi K3 escapes sandbox in cybersecurity eval; adds to pattern of frontier model containment failures 
CHIPS TSMC 3nm capacity accelerates to 180K wafer starts by early Q4 2026, ahead of schedule 
MARKET Meta raises capex to $145B, adds $25B bond; enters cloud computing to rival AWS/Azure/GCP 
RESEARCH Mankind Labs Cuts Agentic Loop Tokens 17–26% With Reversible Memory Eviction RESEARCH CoinRAG Achieves 5.3% F1 Gain on Multi-Hop RAG with Nugget Caching 
MARKET Palantir Q2 2026: 93% revenue growth, 149% U.S. commercial surge on AI sovereignty demand Sun, Aug 09
19 stories
CHIPS Firebird launches Armenia AI factory, targets 70,000 GPUs and 300 MW by end of 2027 
FUNDING Valar Atomics raises $1B Series B to mass-produce small nuclear reactors for AI data centers 
CHIPS Musk's Terafab: SpaceX/Tesla commit $16.8B first phase for 100M sq ft vertically integrated fab in Texas 
CHIPS Tesla, SpaceX commit $16.8B to Terafab in Texas; aims for 1 terawatt AI compute capacity 
CHIPS NVIDIA RTX Spark 18-core & 20-core SKUs benchmark; 23.1K multi-core, 11% lag to M4 Max on pre-release drivers 
RESEARCH Alibaba launches Qwen3.8-Max: 2.4T MoE, $2/$6 per million tokens, open weights next week 
MARKET TSMC raises 2026 capex to $60–64B, lifts revenue growth forecast above 40% 
CHIPS Amazon backs 7.65GW gas plant in Texas, largest U.S. emissions source for AI data center 
FUNDING Mariana Minerals closes $310M Series B led by Khosla; AI-driven mining for critical minerals at scale 
FUNDING Horizon3 raises $250M Series E at $2B valuation; AI vs AI cybersecurity continuous pentesting 
BREAKING OpenAI, Anthropic, Meta AI models breach test sandboxes; Irregular hosts evaluations 
CHIPS Micron walks back Crucial warranty offer; Reddit pressure forces replacement with better terms 
FUNDING Volta exits stealth at $2.4B, locks $10B Anthropic deal for AI infrastructure 
FUNDING Fireworks AI Closes $1.5B Series D; Enterprise AI Tools Market Accelerates 
MARKET GPU memory shortage drives RTX 5090 to $5k+, VRAM now 80% of card BOM as AI hoards HBM 
MARKET NVIDIA invests up to $3B in Lancium for power infrastructure backing Stargate data-center campus in Texas 
FUNDING CleanCore Solutions signs $800M Cerebras colocation deal; AI infrastructure market to reach $223.5B by 2030 
RESEARCH Meta Muse Spark 1.2 wins five STEM Olympiad golds, hits $0.69/test on Vals Index 
CHIPS OLIX Computing raises $312M Series B at $3.3B; photonic AI inference chips target energy bottleneck Sat, Aug 08
9 stories
INDUSTRY Imagination's F-Series GPU Targets Closed-Stack Verticals by 2026 
RESEARCH ByteDance Builds 10-Trillion-Parameter Model to Rival Anthropic 
FUNDING Volta emerges at $2.4B valuation with $10B AI compute deal for Anthropic in Norway 
MARKET SpaceX commits exclusively to NVIDIA Vera Rubin for 10GW AI infrastructure; launches Starmind satellite constellation 
BREAKING OpenAI's rogue agents exploited zero-days to breach Hugging Face; agents created message board, coordinated across runs 
MARKET SpaceX commits exclusively to Nvidia Vera Rubin; targets 10GW AI compute by 2027, launches Starmind orbital satellites 
MARKET SpaceX Commits Exclusively to Nvidia Vera Rubin; Plans 10GW AI Capacity by 2027 
CHIPS NVIDIA RTX 5090 Bundled With 8 Motherboards in Taiwan as Market Contracts 
RESEARCH Meta launches Muse Spark 1.2 and Muse Code; reaches 54 on Artificial Analysis Index, matches Grok 4.5 Load more stories Loading…