Hardware Deals
Monitor AI hardware deals including GPU supply agreements, chip partnerships, and semiconductor contracts with NVIDIA, AMD, Intel, and others.
20
$226.5B
+12 undisclosed
MediaTek reportedly secured an unnamed second AI ASIC customer, with production and revenue contribution expected in 2028.
Applied Materials and Intel agreed to co-develop next generation transistor, interconnect and advanced packaging technologies for AI chips at Applied’s EPIC Center and Intel’s Oregon R&D campus.
SpaceX has reportedly placed a roughly $52 billion order with Foxconn to manufacture over 13,000 Nvidia GB300 AI server racks for its Colossus 1 and 2 data centers.
Collaboration between Mitsubishi Heavy Industries and Nvidia to provide modular, Nvidia DSX-aligned cooling and power infrastructure for large-scale AI data centers.
Intel is investing €5 billion in its Leixlip, Ireland fab to expand production of Intel 3 server chips for AI and high‑performance computing.
Angel++ funding to scale SmartCore’s silicon‑photonics optical I/O chiplets for high‑bandwidth AI data center interconnects.
Series A funding to commercialize OxCore, Oxmiq’s licensable AI GPU architecture aimed at chipmakers that want custom accelerators without full in‑house design cycles.
Halo raised a $7 million seed round to bring its AI-assisted HaloBraid hair braiding device into professional salons globally.
Broadcom and Samsung announced a joint fixed wireless access platform that combines Broadcom’s new Wi‑Fi 8 and 50G PON SoCs with Samsung’s 5G modem to support edge AI workloads.
Samsung Electronics secured Nvidia orders for its first mass-produced HBM4 chips, with customer shipments slated for mid‑February 2026 to power Vera Rubin AI accelerators.
The U.S. Air Force awarded AIM $4.9 million in contracts to deploy its AI platform for fully autonomous heavy machinery in rapid airfield damage recovery.
OpenAI to deploy 6GW of AMD GPUs, AMD issues warrant for up to 160M shares (10% stake)
Lease-to-own deal for AI chips, deploying 300,000-550,000 NVIDIA GB200/GB300 chips for Colossus 2 data center
Nvidia to invest up to $100B in OpenAI for 10GW AI infrastructure deployment starting 2026
Tesla increases reliance on AMD AI processors alongside Nvidia
Samsung supplies AI6 chips for Tesla vehicles and robots
Multi-year GPU supply deal for Tesla's 67k H100/H200 Cortex training cluster
Multi-year supply agreement for H100 and GB200 chips for Azure AI
Multi-year GPU supply for Google Cloud AI infrastructure
Supply agreement for H100 GPUs for AWS AI services
Understanding Hardware deals in AI
Hardware deals are where AI's physical constraints surface: access to GPUs and advanced packaging is the binding limit on how fast models scale, and NVIDIA's dominance makes supply allocation a strategic act. Increasingly these agreements are multi-year capacity commitments rather than spot purchases, locking buyers and suppliers together. Following hardware deals is the clearest read on who can actually train and serve large models at scale.