The U.S. chip giant and Taiwan’s MediaTek say they will work together across AI computing, consumer devices, and automotive technology.
Nvidia (NVDA.US) and Taiwan’s MediaTek (2454.TW) are deepening their partnership to build next-generation AI computing platforms, with Nvidia making its largest-ever direct investment outside the U.S.
The two chipmakers announced the expanded cooperation on Aug. 31, saying they will work together across AI computing, consumer devices and automotive technology.
As part of the deal, Nvidia will invest $3.5 billion in convertible bonds issued by MediaTek. The offering totals $3.9 billion, making it the largest overseas convertible bond issuance in MediaTek’s history. Alphabet is also participating, although the amount of its investment was not disclosed.
Nvidia’s $3.5 billion investment is the company’s largest direct investment outside the U.S. to date.
Three-pronged collaboration
The companies are partnering across three main areas. The most important area is customized AI chips. MediaTek’s custom computing business will gain direct access to Nvidia’s NVLink Fusion interconnect technology, including NVHBM, which Nvidia unveiled last week.
The partnership is aimed at customers who want to develop customized AI processors, or XPUs. Instead of having to engineer and certify every component surrounding a custom chip from scratch, customers can hand the XPU design to MediaTek and focus their resources on the processor itself.
Nvidia and MediaTek can then provide the technology and engineering needed for deployment at scale, including NVLink connectivity, memory architecture, advanced packaging, manufacturing and rack-level systems.
Crucially, custom XPUs developed through the partnership will be able to connect seamlessly to Nvidia’s AI infrastructure.
The expanded partnership will bring the two companies’ supply chains closer together, with Nvidia’s networking ecosystem becoming part of MediaTek’s supply chain and MediaTek’s XPU technology becoming part of Nvidia’s, CEO Jensen Huang told reporters.
The arrangement positions MediaTek to compete more aggressively with Broadcom (AVGO.US) and Marvell Technology (MRVL.US) in the fast-growing market for AI application-specific integrated circuits, or ASICs.
MediaTek expects its AI chip business to generate $2 billion in revenue this year. It is targeting $7 billion to $12 billion in revenue next year, which would give it as much as a 15% share of the global AI ASIC market.
Expanding AI into PCs and cars
The companies will also cooperate on chips designed to deliver AI computing locally on consumer devices, as well as intelligent automotive chips.
Among the few new products to attract significant attention in this year’s consumer electronics market is the N1X chip jointly developed by Nvidia and MediaTek.
Based on Arm (ARM.US) architecture, the N1X combines a 20-core hybrid CPU — 10 high-performance Cortex-X925 cores and 10 efficiency-focused Cortex-A725 cores — with a Blackwell GPU featuring up to 6,144 CUDA cores.
The chip is designed for running AI models locally rather than relying entirely on cloud-based computing. The first PCs equipped with the N1X are reportedly expected to start at close to 20,000 yuan ($2,800), with the first models potentially reaching the market this fall.
The partnership reflects Nvidia’s broader push to extend its accelerated-computing ecosystem beyond the giant data centers that have driven its rise in recent years.
“AI is transforming every computing platform — from the world’s largest AI factories to the PC and the car,” Huang said in a statement. “MediaTek is one of the world’s great semiconductor companies, with exceptional expertise in system-on-chip design, connectivity, leading performance and power efficiency. 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.”
Source:
Cailianshe
