Reports
AI-generated structured vendor updates
TSMC Adds Winbond to WoW 3D Stacking Memory Supply, Breaking DRAM Oligopoly
Winbond joins TSMC's Wafer-on-Wafer (WoW) 3D stacking advanced packaging supply chain, becoming a new DRAM wafer supplier alongside Samsung, SK Hynix, and Micron. This move reduces reliance on the three global DRAM giants and strengthens AI chip packaging supply resilience. Winbond provides DRAM wafers for vertical stacking with TSMC logic wafers, offering 8GB capacity and 256GB/s bandwidth via its CUBE solution.
NVIDIA Space-1 targets orbital AI compute, locking ecosystem with Vera Rubin
NVIDIA hires chief software architect for Space-1, its orbital AI computing system powered by Vera Rubin chips. The system must withstand radiation and temperature extremes. This signals a shift from concept to engineering, though commercial viability remains distant.
Jia Yangqing exits NVIDIA as DGX Lepton shutdown reveals software layer failure
Jia Yangqing leaves NVIDIA after DGX Lepton underperforms and open-source commitments are broken. NVIDIA acquired Lepton AI for ~$700M, rebranded as DGX Cloud Lepton, but service ceased mid-2025. The event signals NVIDIA's failed software layer expansion, shifting control back to hyperscalers.
Samsung and SK Hynix Announce $300B Investment to Dominate AI Memory and Foundry
Samsung and SK Hynix announce a 10-year, 1,000 trillion won investment plan to expand HBM4 production, improve 3nm GAA yield, and build new AI chip fabs. This aims to cement their HBM duopoly and close the gap with TSMC in advanced foundry, reshaping global AI infrastructure supply chain costs.
OpenAI and Broadcom Tape Out First Inference ASIC Jalapeño in 9 Months, Targeting NVIDIA Dominance
OpenAI and Broadcom unveil Jalapeño, their first custom inference ASIC, fabricated on TSMC 3nm and optimized for Transformer models. Targeting a 50% inference cost reduction, it taped out in 9 months and is slated for deployment in gigawatt-scale data centers by late 2026, marking OpenAI's strategic pivot to full-stack AI infrastructure and a direct challenge to NVIDIA's inference hegemony.
Qualcomm Acquires Modular for $3.9B, Open-Sources Mojo to Break CUDA Lock-In
Qualcomm acquires Modular for $3.9B in stock and open-sources Mojo, a Python-compatible systems language. Mojo targets CUDA dependency, aiming to provide a high-performance alternative for AI developers. This move strengthens Qualcomm's AI inference chip software stack and edge AI competitiveness.
NVIDIA Rubin Mandates 100% Liquid Cooling with 45°C Warm Water, Reshaping Data Center Thermal Design
NVIDIA reveals Rubin platform's full liquid cooling design: 100% liquid, 45°C warm water inlet, eliminating chillers and fans. Mass production starts H2 2026, with a mandate for all data centers to transition to liquid cooling, marking a definitive shift in AI thermal management.
Microsoft Cuts Azure China R&D: Geopolitics Forces AI Cloud Retreat
Microsoft is cutting 200-400 Azure R&D roles in Beijing and Shanghai, with departures by July 2026. US AI chip export controls and China's data security laws make frontier AI development impossible. Azure China, operated via 21Vianet, has <5% market share vs Alibaba (30%) and Huawei (19%).
Qualcomm Enters AI Datacenter with Dragonfly ARM CPU, Meta Signs Multi-Generation Deal
Qualcomm unveils Dragonfly C1000 ARM-based datacenter CPU, AI300 accelerator, and interconnect. Meta commits to multi-generation CPU supply, Microsoft Azure to deploy HBC chips. Qualcomm targets $15B+ datacenter revenue by FY2029, acquires Modular for software stack.
OpenAI and Broadcom unveil Jalapeño inference ASIC to bypass NVIDIA GPU dependency
OpenAI and Broadcom launch Jalapeño, a custom ASIC for LLM inference, achieving tape-out in 9 months. OpenAI designs architecture, Broadcom provides networking, Celestica handles integration. Planned for large-scale deployment by end-2026 with gigawatt-scale datacenters, aiming to cut inference costs and reduce NVIDIA dependency.
NVIDIA Unveils Vera CPU for AI Agents, Shifting Control from x86 to Proprietary Silicon
At the annual meeting, Huang announced Vera CPU for AI agents paired with Rubin GPU, claimed Blackwell delivers 30x token throughput over next-best platform, and reiterated CUDA as a moat. This move aims to shift AI compute control from general-purpose CPUs to NVIDIA's proprietary architecture.
TSMC Bets on CoPoS and Glass Substrates: Packaging Paradigm Shifts from Wafer-Level to Panel-Level, AI Chip TCO Inflection
TSMC is replacing CoWoS with CoPoS (panel-level packaging), using 750x620mm square panels and glass core substrates, achieving 20-30% unit area cost reduction. Volume production targets 2028, with AMD Zen 7 as first key customer. This fundamentally alters AI chip packaging economics and capacity scaling.
MediaTek Lands Exclusive Google TPU v9 Inference Upgrade Triggerfish with 2x SRAM
Google plans a TPU v9 inference upgrade, Triggerfish, exclusively fabbed by MediaTek. It features 2-3x on-chip SRAM, HBM4E DRAM, and a simulation die for local management. Production starts late 2027 with 1-2M units lifecycle, unit price ~30% higher than Humufish.
Google TPU v9 Switches to MediaTek, Breaking Broadcom's AI ASIC Monopoly
Google moves its TPU v9 Humufish design and integration contract from Broadcom to MediaTek, which handles I/O chip design and packaging. Combined with a split-foundry strategy (TSMC N2 compute, Samsung 2nm I/O), this marks a systematic effort to build a multi-vendor, multi-node supply chain, directly dismantling Broadcom's dominance in custom AI ASICs.
Intel AI Box Ultra Hits the Road: PC-class Compute Enters Car, Locks Down Edge AI Ecosystem
Intel and Changan Auto launch the AI Box Ultra solution based on the Core Ultra platform, bringing PC-class compute and Android app ecosystem to the cockpit. It emphasizes on-device AI inference, privacy, and offline capability. The move targets Qualcomm and NVIDIA but hides X86 power/thermal drawbacks.
Intel at Computex 2026: CPU as Agentic AI Orchestrator, x86 Reclaims Inference Control
At Computex 2026, Intel unveiled the 288-core Xeon 6+ (Intel 18A) and 3rd-gen Core Ultra, claiming Agentic AI shifts CPU:GPU ratio from 1:8 to 1:1. Partnering with SambaNova and Foxconn for rack-scale inference systems, Intel repositions the CPU as the orchestrator for multi-step AI reasoning, aiming to reclaim control from GPU-centric architectures.
Arm's Self-Designed AGI CPU with Meta: Ecosystem Shift from Licensor to Silicon Vendor
Arm unveils its first self-designed data center CPU, the AGI CPU, with 136 cores on 3nm, purpose-built for agentic AI inference. Co-developed with Meta, which will deploy it across its data centers. Claims 2x rack performance over x86, reducing AI capex by $100B per gigawatt. Signals Arm's shift from IP licensing to direct silicon sales, reshaping ecosystem dynamics.
Apple Expands Private Cloud Compute to Google Cloud with NVIDIA Confidential GPUs
Apple at WWDC 2026 expands Private Cloud Compute (PCC) to Google Cloud, leveraging NVIDIA GPU Confidential Computing for secure AI inference. This marks a strategic shift from Apple-owned data centers to third-party cloud, alongside M6 Neural Engine performance gains.
NVIDIA Launches Arm CPU: RTX Spark and Vera Shift AI Compute Control from x86
NVIDIA unveils RTX Spark Superchip for Windows PC (20 Arm cores, 6144 CUDA, 128GB LPDDR5X) and Vera data center CPU in million-volume production. Vera delivers 1.8x AI workload acceleration over x86. This marks NVIDIA's strategic entry into CPU market, consolidating control via unified Arm+GPU architecture.
Google unveils 8th-gen TPU: 3x training speed, 3x SRAM for inference, redefines AI compute TCO
At Cloud Next 2026, Google launched 8th-gen TPU with dual variants: TPU 8t for training (9600 per pod, 2PB shared memory) and TPU 8i for inference (1152 per pod, 3x on-chip SRAM). Also announced Gemini Enterprise Agent Platform, N4 Axion ARM instances (2x price-performance vs x86), and AI-driven security with Wiz.