Reports
AI-generated structured vendor updates
WhiteFiber and DriveNets Achieve 111.2 Tbps Cross-DC AI Fabric, Breaking Power Constraints
WhiteFiber announces Project Redwood, partnering with DriveNets Ethernet AI fabric (FSE, VOQ, deep buffers), WEKA storage, and NVIDIA H200 GPUs, achieving 111.2 Tbps bandwidth and 0.9ms latency over 83km dark fiber, treating two geographically separated GPU clusters as a single logical supercluster. Commercialization planned for Q3 2026.
Meta Invests $9.17B in Canada AI Data Center, Iris AI Chip Mass Production Begins MTIA Roadmap
Meta announced a $9.17B AI data center in Canada with 1GW capacity, and its first in-house AI chip Iris will mass produce in September, kicking off the MTIA four-generation roadmap. Meta targets 14GW compute by 2027, using 6-month chip iterations to challenge NVIDIA's annual cadence and reduce GPU dependency.
AWS Sells Trainium 3 Externally, Challenging NVIDIA's AI Training Chip Dominance
AWS begins external sales of its Trainium 3 AI training chip, fabricated on TSMC 3nm process, delivering 2.52 PFLOPS per chip. Early customers include Anthropic and Uber. This move directly challenges NVIDIA's dominance and marks AWS's strategic shift from cloud provider to chip vendor.
AMD's Experimental Topological Ghost Protocol Boosts MI300X Inference 10x
AMD introduces experimental Topological Ghost Protocol (TGP) on MI300X GPUs, achieving 431 tokens/sec with 100% success in high-concurrency inference, 10x improvement over standard vLLM. TGP uses KV-cache recycling and segmented state management, still experimental but potentially redefining AI inference benchmarks.
Towards Feature Complete Triton Support in JAX-Triton â ROCm Blogs
...
OpenAI Reopens with GPT-oss Models: Apache 2.0 License Hides Cloud Offload Control
OpenAI launches GPT-oss-120b and GPT-oss-20b under Apache 2.0 license, capable of running on a single 80GB GPU. However, a built-in cloud offload mechanism routes complex queries to proprietary models, masking a strategic control point shift behind the open-source facade.
Google Gemini 3.5 Pro Rebuilds from Scratch: 2M Token Context Window Reshapes AI Frontier
Google DeepMind targets July 17 for Gemini 3.5 Pro, a full architectural rewrite of its pretraining stack to overcome deficits in math reasoning, SVG generation, and image quality. Specs include a 2M token context window, Deep Think reasoning layer, and multi-step autonomous workflows, though unconfirmed by Google.
SambaNova完成11亿美元融资估值110亿美元：推理芯片新格局确立
...
NVIDIA Rigel Core: Single-Threaded CPU as the New Control Plane for Agentic AI
NVIDIA unveils Rosa CPU architecture with custom Rigel core (Arm v9.2), targeting single-threaded performance for Agentic AI workloads, paired with Feynman GPU (1.6nm, 50 PFLOPS) in 2028. This shifts CPU design from core-count scaling to serial-latency optimization, directly challenging AMD EPYC and Intel Xeon dominance.
NVIDIA Vera CPU获Perplexity/OpenAI/Anthropic/Oracle采用 AI Agent性能验证1.5-1.9x加速
...
NVIDIA Vera CPU: Max Single-Threaded Performance at Scale for Agentic AI
NVIDIA launches Vera CPU, a max single-threaded CPU at scale for agentic AI. With Olympus cores delivering 1.8x sustained per-core performance over x86, 1.2TB/s LPDDR5X bandwidth, and 3.4TB/s core-to-core bandwidth, Vera integrates into NVIDIA's unified AI factory architecture, aiming to lock users into its ecosystem.
AI Innovators Adopt NVIDIA Vera — Why Max Single-Threaded CPU at Scale Matters
...
Cisco Locks AI Data Center Security Control Plane with Silicon One and Hypershield
Cisco launches next-gen security for AI data centers, deeply integrating Splunk SIEM with its Silicon One 51.2Tbps chip and Hypershield architecture to push security policies to the network edge. This move aims to shift the security control plane from standalone appliances to its proprietary ASIC and management platform, creating hardware lock-in.
MediaTek and Alibaba Cloud Deploy Tongyi Qianwen LLM on Dimensity Chips
MediaTek partners with Alibaba Cloud to deploy a small version of the Tongyi Qianwen LLM on Dimensity 9300/8300 mobile platforms, enabling offline multi-turn conversations. This move aims to capture edge AI inference control via NPU optimization and SDK integration, directly challenging Qualcomm.
AWS Boosts Trainium3 ASIC Shipments, Accelerating Custom AI Chip Ecosystem Against NVIDIA
Amazon AWS has notified its supply chain to increase Q3 2026 shipments of Trainium3-based ASIC servers by 20-30%. This reflects growing confidence in its custom AI chips and a strategic push to reduce reliance on NVIDIA GPUs. AWS also partnered with OpenAI to develop a Stateful Runtime Environment on Bedrock.
NVIDIA Denies Kyber NVL144 Delay, But 78-Layer PCB Bottleneck Exposes AI Hardware Physics Limit
NVIDIA officially denies reports of Kyber NVL144 rack delay to 2028, but SemiAnalysis revelations about a 78-layer ultra-high-density PCB midplane bottleneck and Rubin Ultra cancellation expose hard physical limits in signal integrity and manufacturing, opening a strategic window for AMD and Google.
AWS boosts Trainium 3 shipments, accelerating ASIC substitution for NVIDIA GPUs
Supply chain sources indicate Amazon AWS has instructed vendors to increase Trainium 3 shipments for Q3 2026 by 20-30%. This signals strong confidence in its custom ASIC strategy to reduce dependence on NVIDIA GPUs, leveraging superior cost and power efficiency for cloud AI training.
NVIDIA Kyber NVL144 Delayed to 2028: Midplane PCB Manufacturing Becomes AI Scaling Bottleneck
SemiAnalysis reveals NVIDIA's Kyber NVL144 delayed beyond 12 months to 2028 due to 78-layer Orthogonal Backplane manufacturing challenges. The interim NVL72x2 solution is cancelled due to operational burdens, and the 4-die Rubin Ultra is also scrapped, leaving a product gap in NVIDIA's scaling roadmap.
Anthropic Starts Custom AI Chip Development, Talks Samsung 2nm, Aims for Compute Independence
Anthropic has initiated its own AI chip development and is in talks with Samsung for 2nm foundry services. The move aims to reduce reliance on NVIDIA GPUs, optimize inference costs, and strengthen its technology moat ahead of a potential IPO. It joins OpenAI, Google, and others in the custom ASIC race, signaling a shift from software to hardware competition.
AMD Unveils Zen 6/7 CPU and MI400/500 GPU Roadmap, Targets NVIDIA Rubin with HBM4 and 2nm
AMD unveiled its Zen 6/7 CPU and MI400/500 GPU roadmap at its 2026 Financial Analyst Day, featuring TSMC 2nm process and HBM4 memory. The MI400 series boasts 432GB memory, 19.6TB/s bandwidth, and 40 PFLOPs FP4 performance, directly targeting NVIDIA's Vera Rubin architecture with an annual cadence to disrupt the AI hardware monopoly.