Qualcomm (QCOM-US) announced on Tuesday (8th) a partnership with Amazon (AMZN-US) to co-develop AI data center chips. The two companies will jointly develop customized chips across multiple product generations to help Amazon Web Services (AWS) expand its AI inference-focused infrastructure. This agreement is seen as a major breakthrough for Qualcomm's entry into the data center market, driving its stock price up 9.7% in pre-market trading and maintaining a gain of over 4% at market open.
At the time of reporting, Qualcomm (QCOM-US) shares rose 4.10%, temporarily trading at $175.66 per share.
According to regulatory filings, Qualcomm has issued warrants to Amazon, granting it the right to purchase up to 25 million shares of Qualcomm stock at $161.26 per share, representing a potential equity value of approximately $4 billion. The warrants will vest in tranches based on the order value Amazon places with Qualcomm, with total orders potentially reaching up to $60 billion.
The companies did not disclose other financial terms of the agreement. In addition to AI inference chips, Qualcomm and Amazon will jointly develop high-speed optical connectivity technology capable of transmission speeds up to 1.6 terabits per second, addressing the rapidly growing bandwidth demands of AI data centers. Qualcomm also plans to expand its use of AWS's AI services and infrastructure for chip design work, aiming to shorten product development cycles.
Qualcomm is best known for smartphone and mobile device processors but has been actively expanding its AI data center business to reduce reliance on the mobile market. In June, the company launched the Dragonfly C1000 CPU, specifically designed for data centers, targeting agent-based AI workloads and energy-efficient computing, and announced that Meta Platforms (META-US) will adopt it after mass production begins in 2028.
Qualcomm aims to generate $15 billion in data center revenue by fiscal year 2029 and plans to launch AI accelerator chips and interconnect products that link multiple chips. With support now from Amazon, another hyperscale cloud provider following Meta, market confidence in Qualcomm's data center strategy has strengthened. Both Meta and Amazon invest tens of billions of dollars annually in AI infrastructure capital expenditures.
The AI chip market is currently dominated by NVIDIA (NVDA-US), whose GPUs excel at parallel processing and are widely used for training and running AI models. However, as agent-based AI grows rapidly, CPUs responsible for general-purpose computing are becoming increasingly important.
Bank of America (BAC-US) predicts the global CPU market could more than double from $27 billion in 2025 to $60 billion by 2030. Demand for data center CPUs from Intel (INTC-US) and AMD (AMD-US) is growing rapidly, and NVIDIA has already launched CPUs optimized for agent-based AI. NVIDIA's AI infrastructure head, Dion Harris, previously noted that CPUs are increasingly becoming a bottleneck in scaling AI and agent-based workflows.
FACT BOX
- Source: PR Times
- Category: Partnership
- Organizations: Amazon / Meta Platforms / NVIDIA
- Products / services: Dragonfly C1000 CPU