Reports
AI-generated structured vendor updates
AMD Mustang Peak Threadripper: 144 cores, PCIe 6.0, TR6 socket – Power and memory challenges loom
AMD's Zen 6 Threadripper 'Mustang Peak' is confirmed with 2nm TSMC process, DDR5, PCIe 6.0, and a new TR6 socket. Using Powderhorn CCDs, it scales to 144 cores (288 threads) with clocks above 6 GHz. However, massive power draw and memory bandwidth demands (possibly requiring MRDIMM) raise platform cost concerns.
Meta面临欧盟巨额罚款风险WhatsApp AI功能受阻
...
NVIDIA RTX Remix 1.5: RTX IO Shrinks Game Sizes, AI Agents Reshape Modding
NVIDIA releases RTX Remix 1.5, featuring RTX IO compression that slashes Half-Life 2 RTX from 80GB to 50GB and reduces CPU overhead. The update also introduces AI agent integration via 'RTX Remix Skills,' allowing AI coding agents to automate complex modding tasks, lowering the barrier for non-programmers.
NVIDIA & Coherent Expand 6-Inch InP Fab, Locking AI Optical Interconnect Supply Chain
Coherent breaks ground on the world's first 6-inch indium phosphide fab in Texas, backed by $2B from NVIDIA and multi-billion purchase commitments. The facility produces lasers, transceivers, and pluggable optics for silicon photonics interconnects, enabling NVIDIA's Vera Rubin Ultra NVL576 576-GPU clusters and signaling a mass shift from copper to optical backbones in AI data centers.
Salesforce Acquires Fin for $3.6B: AI Agent Ecosystem Consolidates into CRM
Salesforce acquires Fin (formerly Intercom) for $3.6B to bolster its Agentforce platform. Fin brings 30,000 customers, a 76% autonomous resolution rate, and proprietary model Apex. The move fills a gap in customer service automation, accelerates CRM+AI agent bundling, and sets a valuation benchmark for the AI agent market.
Lexar Offloads AI Models to SSD: DRAM Cut 40%, Latency Remains Hurdle
Lexar unveils AI Storage Core SSD with a custom SPU DRAM-less controller and software stack, offloading LLMs to NAND Flash. It runs Qwen 3.5 122B on 32GB DRAM at 15.6 tokens/s (3x improvement), but TTFM latency of 2-8 seconds hinders real-time use.
SiMa.ai Palette Neat: Natural-Language Agentic Environment Dismantles NVIDIA's GPU Moat
SiMa.ai launches open-source Palette Neat, an agentic development environment for Physical AI, paired with its sub-10W Modalix SoM. It uses natural language to abstract compute complexity, slashing dev cycles from months to days. Pin-compatible with NVIDIA SoM, it targets breaking the GPU ecosystem lock-in.
AMD Ryzen 10000 Series to Swap iGPU for NPU: AI Boost at Cost of Basic Display
Leaks suggest AMD's next-gen Zen 6 desktop CPU 'Olympic Ridge' will replace the integrated GPU with an NPU, targeting >40 TOPS for Copilot+ AI PC certification. It also upgrades the client I/O die to support CUDIMM/CAMM and EXPO 1.2 for faster DDR5. The trade-off boosts local AI but forces nearly all users to rely on a discrete GPU for basic display.
ASML, TSMC, imec Demo 300mm 2D-Material Transistors at 50nm Pitch
imec, ASML, and TSMC demonstrate the first 300mm wafer integration of MoS2/WS2/WSe2-based n and pFETs with 50nm contacted poly pitch (CPP) using single-patterning EUV lithography, achieving 94% operational yield. This lab-to-fab breakthrough paves the way for 2D channel materials to extend Moore's Law.
NVIDIA's Desktop DGX Station with GB300 Shifts Control from Cloud to Local Hardware
ASUS launches ExpertCenter Pro ET900N G3, built on NVIDIA DGX Station GB300 architecture with GB300 Grace Blackwell Ultra chip, 748GB coherent memory, and 20 PFLOPS AI performance. This deskside AI supercomputer enables local LLM fine-tuning, inference, and agentic AI workflows via NVLink-C2C and the full NVIDIA AI software stack including NemoClaw.
Google Cloud Next '26: Agent Gateway Seizes Control Plane, TPU 8i Locks Inference
Google Cloud Next '26 announces 8th-gen TPUs (8t for training, 8i for inference), Agent Platform with Agent Gateway, Agent Identity, Agent-to-Agent Orchestration, Agentic Data Cloud, and Agentic Defense integrating Wiz. The move shifts control from infrastructure to agent orchestration, locking enterprises into a vertically integrated stack.
Meta Unveils Muse Spark Foundational Model and Re-architects AI Assistant
Meta launched Muse Spark, the first model from its Superintelligence Labs, using it to overhaul the Meta AI assistant. The new architecture enables parallel subagents for reasoning, robust multimodal perception, and leverages social graph content for personalized responses.
Meta WhatsApp Launches Cross-Platform Migration and AI Features
Meta WhatsApp introduced four feature updates: cross-platform chat migration, dual-account support on iOS, storage management, and AI photo editing. These focus on consumer experience optimization without enterprise technical architecture changes.
Meta Launches AI Anti-Fraud Platform with Multi-Modal Analysis and Verification
Meta deploys an AI-powered multi-modal anti-fraud system combining text, image, and contextual analysis for enhanced scam detection. The platform expands to WhatsApp, Facebook, and Messenger with new warning features and strengthens advertiser verification processes.
Meta Collaborates with Law Enforcement and Launches Anti-Scam Tools
Meta partnered with global law enforcement to disrupt scam networks, disabling 150K accounts and making arrests. It launched three anti-scam tools including suspicious request alerts and AI-powered fraud detection, enhancing protection through proactive warnings and real-time information sharing.
Meta Launches AI Anti-Fraud Tools and Enhances Global Law Enforcement Collaboration
Meta adopts advanced AI to analyze text, image, and contextual signals for detecting complex fraud patterns. New anti-fraud tools include suspicious friend request warnings and chat scam detection. Expands collaboration with law enforcement and enhances ad verification.
Vonage Launches Omnichannel Conversations for Marketing Platform Enhancement
Vonage introduces an omnichannel conversations solution, integrating multi-channel customer engagement via SMS, WhatsApp, etc. It features a low-code dialog builder, real-time analytics dashboard, and API integration with CRM systems, expanding its customer interaction portfolio for end-to-end communication.
Google Launches Festival AI Music Generation Feature Based on Lyria 3 Model
Google introduced a limited-time AI music creation feature in Gemini app, using Lyria 3 model to generate personalized music cards. The feature allows multi-region users to create 30-second hi-fi audio and custom covers through prompts, with social sharing integration.
Meta Introduces New Alerts for Instagram Parental Supervision to Flag Teen Suicide/Self-Harm Searches
Meta announced that it will roll out a new safety alert feature for its Instagram parental supervision tools in the coming weeks. The system will automatically notify parents or guardians if a supervised teen account repeatedly attempts to search for content related to suicide or self-harm within a short period. Alerts will be sent via email, text, WhatsApp, or in-app notifications and will include expert resources to guide parents in having sensitive conversations with their teens. The feature was developed based on analysis of Instagram search behavior and consultation with experts from Meta's Suicide and Self-Harm Advisory Group. It employs a specific threshold (multiple searches in a short time) to balance effective warning with avoiding over-notification. The rollout will begin next week in the US, UK, Australia, and Canada, with expansion to other regions planned later this year. Furthermore, Meta revealed it is building similar parental alerts for certain AI interactions. Notifications will be sent if a teen attempts to engage in specific types of conversations related to suicide or self-harm with Meta's AI. More details will be shared in the coming months. **Comment**: This represents another concrete product feature from Meta in the realm of youth online safety, combining content blocking with parental intervention to find a balance between platform responsibility and family support. Its technical implementation relies on precise content identification and threshold setting, but its effectiveness hinges on alert accuracy and subsequent parental action. Companies can observe this "platform-family"联动 safety model for its applications in compliance and product design.
Meta and RAI Release AI-Driven Retail Whitepaper
Meta partnered with RAI to release a whitepaper on AI-driven omnichannel retail solutions, highlighting social platforms as key discovery channels and WhatsApp's rapid growth as a commerce tool.