Qualcomm and Amazon Announce AI Data-Center Deal

Qualcomm and Amazon have announced custom AI-inference silicon and optical links up to 1.6 Tbps. Further warrant vesting depends on commercial arrangements, binding orders and purchases.

TL;DR
  • Deal: Qualcomm and Amazon agreed to develop customized AI-inference silicon and optical links supporting speeds up to 1.6 terabits per second.
  • Warrant: An Amazon affiliate may buy up to 25 million Qualcomm shares as commercial milestones and purchases vest tranches.
  • Limits: Neither company disclosed product architecture, quantities, delivery dates, measured performance or ordinary financial terms.

Qualcomm and Amazon announced a multi-generation data-center collaboration on September 8 that covers customized silicon for AI inference and optical links reaching up to 1.6 terabits per second. A separate purchase-linked warrant, already partly vested on initial purchase commitments, gives the engineering plan unusual commercial weight even though neither company disclosed products, delivery dates or ordinary financial terms.

The agreement puts Qualcomm’s data-center expansion inside a major cloud operator that already builds its own processors and AI accelerators. It also joins two complementary constraints: chips must run trained AI models, while high-speed connections must move data among the machines doing that work. The warrant makes Amazon’s future purchases relevant to both Qualcomm’s sales ambitions and possible future Amazon ownership of Qualcomm shares.

What the Two Technical Tracks Cover

 Inference is the stage at which a trained AI model processes new input and generates an answer or prediction. Customized silicon means chips tailored to a customer’s workloads and systems.

The companies did not say whether the resulting designs will be CPUs, GPUs or another type of accelerator. A CPU handles a broad range of general-purpose and control work. GPUs perform many calculations in parallel, which makes them widely used for AI training and inference. Specialized accelerators can take on narrower AI tasks, and all three types can operate in the same data center.

Qualcomm will contribute serializer-deserializer, or SerDes, technology that turns parallel chip data into high-speed serial streams and back again, plus optical digital signal processors that condition those streams for transmission over light. The announced ceiling of 1.6T refers to roughly 1.6 terabits of data per second for a link or transceiver class. 

Qualcomm also plans to use more AWS AI infrastructure, including Amazon Bedrock, for its chip-design work and to shorten design cycles.

Commercial Milestones Can Vest Warrant Tranches

A September 8 SEC filing separates the commercial actors behind the agreement: Qualcomm Incorporated issued the warrant; Qualcomm Technologies and certain affiliates are the supplier side; Amazon Data Services and certain affiliates are the purchaser side; and Amazon.com NV Investment Holdings holds the warrant.

The warrant gives the Amazon investment affiliate the right to acquire as many as 25 million Qualcomm common shares at an initial exercise price of $161.26 per share. The Amazon affiliate can exercise vested rights without paying cash, and the warrant expires on September 3, 2036.

The affiliate holds rights to acquire shares rather than shares it already owns. A tranche covering 3.75 million warrant shares vested at issuance because of initial purchase commitments.

More tranches can vest as the companies enter specified commercial arrangements, the Amazon purchaser group places binding orders and that group makes actual purchases. Eligible purchases include Qualcomm server-chip products, technology, systems and manufacturing services. The filing caps the payments counted by the vesting schedule at $60 billion, rather than committing Amazon to spend that amount or guaranteeing Qualcomm that revenue.

Until exercise, the warrant gives its holder no Qualcomm shareholder voting rights. Future exercise could lead to share issuance, but cashless exercise, conditional vesting and customary adjustments prevent the share cap and strike price from defining cash already invested or the value of a present Amazon stake.

Amazon Adds Weight to Qualcomm’s Data-Center Expansion

The September agreement advances a strategy Qualcomm had already made public. Its June Dragonfly roadmap grouped data-center CPUs, AI accelerators, custom silicon and connectivity into one portfolio. Qualcomm also named Meta as a future customer for its Dragonfly C1000 server CPU and set a target of more than $15 billion in data-center revenue in fiscal 2029. Those plans establish the baseline for September; the Amazon collaboration advances the roadmap without yet producing a disclosed product or revenue.

AWS already designs Nitro infrastructure chips, Graviton CPUs, Inferentia inference accelerators and Trainium training accelerators. Qualcomm is therefore seeking a place within an established mix of AWS-designed and third-party hardware, not filling an empty product category.

Qualcomm is now clearly moving beyond the mobile processors that built its business into data-center compute and interconnects, while Nvidia remains the dominant supplier of AI accelerators in the partnership. For now, the collaboration creates an opportunity inside AWS’s hardware mix without establishing a shipped Qualcomm product or displacement of Nvidia or AWS-designed chips.

The purchase-linked warrant also turns later commercial activity into a potential equity relationship. If the Amazon purchaser group enters the specified arrangements and buys eligible Qualcomm products or services, more warrant tranches can vest; if the separate investment affiliate exercises vested rights, Qualcomm can issue shares under the warrant’s terms. Any larger competitive effect will depend on Qualcomm turning the two workstreams into named products and Amazon placing the orders that release those rights.

Markus Kasanmascheff
Markus Kasanmascheff
Markus has been covering the tech industry for more than 15 years. He is holding a Master´s degree in International Economics and is the founder and managing editor of Winbuzzer.com.
Subscribe
Notify of
guest
0 Comments
Newest
Oldest Most Voted