Back to News
IndustryAI Understanding briefing

Qualcomm signs multigenerational AI chip deal with Amazon

Qualcomm has agreed to supply Amazon Web Services with custom AI data center chips and optical connectivity, a move that validates its infrastructure push but does not immediately threaten Nvidia's dominance.

4 min readRead the primary source
Source-provided image accompanying Qualcomm signs multigenerational AI chip deal with Amazon
Source referenceSource recorded
Publisher
finance.biggo.com
Source link
finance.biggo.comhttps://finance.biggo.com/news/12a00204-c55d-4c35-8179-53d068dc5b2c
Source type
Linked source — primary-source status has not been established.

Story last revised

ContextUnderstand this in 60 seconds

Start here

Key terms

Inference
The runtime phase where a trained model generates predictions or outputs.
Test yourselfAI Models Explained Quiz

What happened

Qualcomm announced a multigenerational agreement to provide custom AI data center chips and optical connectivity to Amazon Web Services. The deal includes warrants for Amazon to acquire 25 million Qualcomm shares at $161.26, tied to up to $60 billion in chip purchases over ten years, with revenue expected to begin in the December quarter.

Qualcomm has locked in a multigenerational agreement to supply Amazon Web Services with custom AI data center chips and optical connectivity. The partnership, announced on September 9, sent Qualcomm shares up 3% as investors digested the terms. According to a securities filing, Amazon received warrants to acquire 25 million shares at $161.26 apiece, representing a potential $4 billion equity stake. These warrants expire on September 3, 2036, and vest in tranches tied to commercial milestones and purchases of up to $60 billion in Qualcomm server chips and related technology over a 10-year horizon.

The collaboration focuses heavily on inference workloads, which involve running AI models after they have been trained. Qualcomm stated that the effort spans multiple generations of custom silicon aimed at AWS's expanding AI infrastructure, with an emphasis on energy-efficient processing and system-level integration. Revenue from this arrangement is expected to begin in the December quarter. This announcement represents a second major hyperscaler endorsement for Qualcomm in less than three months, following the June unveiling of the Dragonfly C1000 data center CPU, which Meta Platforms is set to adopt when volume production begins in 2028.

Bernstein analyst Stacy Rasgon characterized the deal as confirmation of Qualcomm's existing trajectory rather than evidence of upside beyond current forecasts. After speaking with Qualcomm's investor relations team, Rasgon concluded the agreement does not appear incremental to the company's $5 billion fiscal 2027 and $15 billion fiscal 2029 revenue targets. He noted that the deal confirms Amazon is one of Qualcomm's two hyperscaler customers and that initial purchase commitments will trigger 3.75 million warrants, which he estimates could be worth approximately $9 billion if they vest at a steady pace.

Source details: finance.biggo.com

Why it matters

The partnership confirms Qualcomm's entry into the high-stakes AI infrastructure market with a major hyperscaler, providing a long-term revenue runway and validating its custom silicon strategy. However, analysts note the deal focuses on inference workloads and does not challenge Nvidia's GPU dominance in training, positioning Qualcomm as a credible but secondary player in the AI chip hierarchy.

The broader question for investors is whether Qualcomm's data center push can meaningfully erode Nvidia's commanding position in AI silicon. Based on the structure of the Amazon deal, the answer appears to be no in the near term. Nvidia's dominance rests on its GPU architecture, which excels at both training and running AI models due to parallel processing capabilities. In contrast, Qualcomm's agreement focuses on custom CPUs and inference-focused silicon, a meaningful but secondary position in the AI infrastructure stack.

Bank of America projects the CPU market could more than double from $27 billion in 2025 to $60 billion by 2030, driven by demand from various vendors. However, Qualcomm's $15 billion data center revenue target for fiscal 2029 pales in comparison to Broadcom's expected $115 billion in AI revenue for fiscal 2027 and Nvidia's annualized data center revenue exceeding $200 billion. The market's muted reaction, with shares rising just 3%, suggests investors understand that Qualcomm is carving out a niche rather than disrupting the established competitive hierarchy.

This deal validates Qualcomm's push into the infrastructure market and aligns Amazon's interests with Qualcomm's success through the warrant structure. While it does not reshape the competitive landscape dominated by Nvidia, it provides Qualcomm with a long runway and two anchor customers in Meta and Amazon, solidifying its status as a credible entrant in the AI chip market.

What to watch next

Investors should monitor Qualcomm's ability to meet its $15 billion fiscal 2029 data center revenue target and the pace of warrant vesting tied to commercial milestones. Additionally, the competitive dynamics between Qualcomm, Broadcom, and Nvidia in the custom AI accelerator space will determine long-term market share shifts.

Investors should monitor Qualcomm's ability to achieve its $15 billion data center revenue target for fiscal 2029, a figure management now says it has 'very high confidence' in achieving partly because of the Amazon agreement. The pace of warrant vesting tied to commercial milestones and chip purchases will also be a key indicator of the deal's progress.

The competitive dynamics between Qualcomm, Broadcom, and Nvidia in the custom AI accelerator space will determine long-term market share shifts. Broadcom has already established itself as the leading provider of custom AI accelerators, while Nvidia remains the undisputed leader in AI silicon. Qualcomm's success in replicating these relationships and design expertise will be crucial for its long-term viability in the data center market.

Related guides & quizzes

AI Models ExplainedFuture of AIAI TrainingTest what you know — try a free AI quizLook up an AI term in our glossary
Found this useful?