Nvidia $3.5B investment in MediaTek counters big-tech AI chip buildout
In another example of what some criticize as “circular financing,” Nvidia is investing $3.5 billion in convertible bonds issued by MediaTek, with MediaTek adopting the Nvidia NVLink Fusion platform and collaborating on RTX Spark™ and DGX Spark™ PC chips. Some analysts say this gives Nvidia a competitive edge, while others argue “circular financing” adds further risk.
Jon Markman for Forbes says the deal makes Nvidia a “tollbooth for the custom chip era,” with a structural defense mechanism that ensures Nvidia’s architecture and plumbing will be needed to plug into data center racks, even when companies choose to design their own specialized AI processors.
Nvidia’s Jensen Huang in his announcement said, “together, we’re building platforms that bring NVIDIA accelerated computing to new markets and give customers the freedom to create differentiated AI systems at enormous scale.”
Rick Tsai, vice chairman and CEO of MediaTek said “by combining NVIDIA’s leadership in accelerated computing and AI software ecosystem with MediaTek’s expertise in a diverse AI technology portfolio from edge to cloud, and our leadership position in custom silicon, we can accelerate innovation for our customers.”
Indeed, with this move, Nvidia goes beyond just GPUs and extends its reach into the AI infrastructure ecosystem, which understandably, makes some on Wall Street critical of vendor-financed rather than organic growth. But still there are others that tout this as as a masterstroke that turns the shift toward custom AI processors into a major win for Nvidia’s bottom line.
Susana Schwartz
Technology Editor
RCRTech
AI Infrastructure Top Stories
SKT launches SK Horizon: SK Telecom has signed an agreement with funds managed by KKR and the IMM Investment-Stonebridge consortium for a KRW 3.08 trillion ($2.1 billion) equity investment in SK Horizon’s 318 MW of capacity.
OpenAI Jalapeño performance: Jalapeño’s reticle-sized design brings 13.4 petaFLOPS of mixed-precision matrix compute using mxfp4 x mxfp4 (per accelerator) – lowering latency by 1.7x compared to an Nvidia GB200.
AI Today: What You Need to Know
SLB adds Kelvion to its portfolio: SLB will acquire thermal management company Kelvion, extending its Data Center Solutions business, whose revenue is projected to grow at a 90% CAGR, delivering capacity in excess of 2 GWs by end of year..
AMD, Cisco, Humain in Saudi Arabia: AMD, Cisco, and Humain will jointly deploy production AI compute for AI model development, with AMD Instinct MI355X graphics processors, EPYC central processors, and Cisco networking equipment.
AWS $5B investment in Saudi Arabia: AWS announced its 1st cloud region in Saudi Arabia alongside an AI Zone commitment with Humain. Jointly, they plan to deliver 50MW of AI capacity using AWS Trainium chips and NVIDIA infrastructure.
Broadcom private AI: Broadcom revealed VMware Private AI Cloud, which gives enterprises the ability to run AI production on premises, and VMware AI Factory, which serves as the software-defined foundation of VMware Private AI Cloud.
OpenAI’s purchase of Apple hardware: OpenAI has purchased tens-of-thousands of Mac minis and Mac Studios to train computer-use agents, with thousands of independent machines beating raw GPU horsepower for task-completion training.
RCR Events
Industry Resources
Report: NTN in motion — evolving standards, expanding services