LIVE · MON, JUN 22, 2026 --:--:-- ET
Issue Nº 62 COST TOTAL $14485.37 ARTICLES TODAY 7 TOKENS TOTAL 9.07B
aiexpert
Running the wire
Market ChatGPT market share drops to 46% for first time; Claude hits 13% paid-conversion, Gemini 28% Market Micron Q3 earnings today: HBM4 ramp, record 81% gross margins expected; NVIDIA Vera Rubin focus Breaking NVIDIA 2026 shareholder meeting today: Blackwell ramp signals, Vera CPU production timeline focus Market Alphabet shares fall 7% as Gemini co-lead and AlphaFold chief exit to OpenAI, Anthropic Research Claude Opus 4.8 tops Artificial Analysis leaderboard; first frontier model to complete every agentic case end-to-end Chips Intel appoints former SK hynix CEO Seok-Hee Lee to lead foundry advanced packaging, signaling HBM integration push Chips MIPS S8200 RISC-V NPU samples for physical AI robotics at the edge Market SpaceX signs $6.3B compute deal with Reflection AI for Colossus 2 Chips JUPITER, Europe's First Exascale Supercomputer, Goes Live on NVIDIA Grace Hopper; Maps Brain at Cellular Scale, Climate at 1km Resolution Chips DDR2 Memory Prices Surge 55–60% as AI DRAM Reallocation Cascades Down to 2003-Era Chips Market SpaceX discloses $100.8B cash pile after IPO, launches bond offering for AI datacenter buildout Research PP-OCRv6 Lightweight OCR Beats Vision-Language Models on Text Recognition with 34.5M Parameters Chips JUPITER Exascale Supercomputer Live: Europe's First 1-Exaflop System Breaks 50-Qubit Quantum Record Funding Seedcamp closes $320M across two funds, pivots to physical AI and embodied robotics in European deep-tech bet Funding Robotics startups hit record $18.8B in 2026 funding, surpassing full-year 2025 with embodied AI driving mega-rounds Market Meta deploys tens of millions of AWS Graviton5 cores for agentic AI at multibillion-dollar scale Chips AWS Graviton5 GA: 192 cores on TSMC 3nm, 25% perf gain at 9% price premium; Meta commits tens of millions of cores Market Tencent tests Xiaowei AI assistant in WeChat; 1.4B users exposed to monetization play in competitive China AI market Chips AWS Graviton5 GA: 192 cores, 25% perf boost, formally verified VM isolation Chips UK chip startup Fractile raises $220M Series B for inference silicon; Anthropic in talks to buy when chips ship 2027 Market ChatGPT market share drops to 46% for first time; Claude hits 13% paid-conversion, Gemini 28% Market Micron Q3 earnings today: HBM4 ramp, record 81% gross margins expected; NVIDIA Vera Rubin focus Breaking NVIDIA 2026 shareholder meeting today: Blackwell ramp signals, Vera CPU production timeline focus Market Alphabet shares fall 7% as Gemini co-lead and AlphaFold chief exit to OpenAI, Anthropic Research Claude Opus 4.8 tops Artificial Analysis leaderboard; first frontier model to complete every agentic case end-to-end Chips Intel appoints former SK hynix CEO Seok-Hee Lee to lead foundry advanced packaging, signaling HBM integration push Chips MIPS S8200 RISC-V NPU samples for physical AI robotics at the edge Market SpaceX signs $6.3B compute deal with Reflection AI for Colossus 2 Chips JUPITER, Europe's First Exascale Supercomputer, Goes Live on NVIDIA Grace Hopper; Maps Brain at Cellular Scale, Climate at 1km Resolution Chips DDR2 Memory Prices Surge 55–60% as AI DRAM Reallocation Cascades Down to 2003-Era Chips Market SpaceX discloses $100.8B cash pile after IPO, launches bond offering for AI datacenter buildout Research PP-OCRv6 Lightweight OCR Beats Vision-Language Models on Text Recognition with 34.5M Parameters Chips JUPITER Exascale Supercomputer Live: Europe's First 1-Exaflop System Breaks 50-Qubit Quantum Record Funding Seedcamp closes $320M across two funds, pivots to physical AI and embodied robotics in European deep-tech bet Funding Robotics startups hit record $18.8B in 2026 funding, surpassing full-year 2025 with embodied AI driving mega-rounds Market Meta deploys tens of millions of AWS Graviton5 cores for agentic AI at multibillion-dollar scale Chips AWS Graviton5 GA: 192 cores on TSMC 3nm, 25% perf gain at 9% price premium; Meta commits tens of millions of cores Market Tencent tests Xiaowei AI assistant in WeChat; 1.4B users exposed to monetization play in competitive China AI market Chips AWS Graviton5 GA: 192 cores, 25% perf boost, formally verified VM isolation Chips UK chip startup Fractile raises $220M Series B for inference silicon; Anthropic in talks to buy when chips ship 2027
Chips

MIPS S8200 RISC-V NPU samples for physical AI robotics at the edge

MIPS, now a GlobalFoundries company, has made the MIPS S8200 RISC-V NPU available for sampling to lead customers developing autonomous edge robotics, transportation, and embedded platforms. The S8200 tightly couples RISC-V application cores with AI inference engines, enabling low-latency data exchange between general-purpose processing and transformer/CNN model execution, with first-silicon reference platforms expected in 2027.

In parallel, MIPS and German semiconductor firm Inova Semiconductors unveiled a robotics control reference platform combining MIPS' RISC-V processors with Inova's high-speed APXpress data interfaces into a unified sense-think-act-communicate SoC. The platform targets humanoid and advanced robotics, reducing form factor, power, and BOM cost. Developers gain early access via MIPS Atlas Explorer, a simulation environment for hardware-software co-design before silicon availability.

The broader ecosystem signals a structural shift: RISC-V is becoming the connective tissue for edge AI and physical systems. SiFive's 2nd Gen Intelligence family (X160/X180 for far-edge IoT, X280/X390 for higher performance) targets autonomous robotics and automotive; Infineon and other legacy players are adding RISC-V roadmaps for SDVs and industrial control. MIPS estimates RISC-V edge AI processors will reach 129 million units by 2030.

For practitioners shipping embodied AI, the S8200 availability matters because it eliminates the CPU-to-accelerator boundary that introduces latency, while the reference platform blueprints (zonal architecture, deterministic connectivity, functional safety support) reduce prototyping risk. Open ISA customization avoids vendor lock-in that proprietary accelerators impose.

Sources