Qualcomm Technologies has announced a multi-generational collaboration with Amazon focused on customized silicon for large-scale AI data centers, as the companies expand their work across AI inference and advanced optical connectivity.
The collaboration will span multiple generations of customized silicon designed to support the continued growth of Amazon Web Services’ AI infrastructure.
Qualcomm and Amazon are also working together on optical connectivity technologies designed to support increasingly bandwidth-intensive AI workloads, with solutions extending to 1.6T and future generations.
The agreement comes as AI workloads drive significant growth in demand for compute capacity, storage, networking, memory bandwidth and energy-efficient infrastructure.
Qualcomm said the collaboration combines Amazon’s AI infrastructure capabilities with its own expertise in power-efficient processing, semiconductor design and system-level integration.
A major component of the partnership will involve AI inference, where specialized chips process trained AI models and generate outputs for users and applications.
The companies are also targeting the networking bottlenecks created by increasingly large AI clusters.
Qualcomm Technologies and Amazon plan to collaborate on high-performance optical connectivity products that can move large volumes of data between systems inside AI data centers.
The networking work will use Qualcomm’s advanced SerDes and optical digital signal processor technologies.
SerDes technology converts data between serial and parallel formats and is used to support high-speed communication between semiconductor components and infrastructure.
As AI systems grow in scale, those interconnects are becoming increasingly important because performance depends not only on individual processors but also on how quickly data can move between accelerators, memory, storage and networking equipment.
The partnership represents a further expansion of Qualcomm’s ambitions beyond its traditional mobile and edge-computing businesses and into large-scale AI data center infrastructure.
Qualcomm has increasingly positioned its power-efficient semiconductor architecture and connectivity expertise as applicable across computing environments ranging from personal devices to hyperscale infrastructure.
The Amazon collaboration gives the company an opportunity to apply those capabilities to customized silicon specifically designed around the requirements of AWS AI infrastructure.
The companies did not disclose specific product names, production timelines or financial terms for the customized silicon collaboration.
However, describing the agreement as multi-generational signals that the partnership is intended to extend beyond a single semiconductor program.
The relationship also includes a deeper use of AWS technology within Qualcomm’s own chip development processes.
Qualcomm Technologies plans to increase its use of AWS AI infrastructure, including Amazon Bedrock, for electronic design automation workloads.
The company is targeting shorter semiconductor design cycles by applying AI infrastructure to parts of the EDA process.
Electronic design automation tools are used throughout semiconductor development to design, simulate, verify and optimize increasingly complex chips.
Applying AI systems to those workloads could potentially help engineers automate or accelerate parts of the development process, which is becoming more computationally demanding as semiconductor architectures grow more sophisticated.
For AWS, the collaboration adds another semiconductor partner as it continues investing in customized infrastructure designed around AI workloads.
The partnership is focused not only on computing silicon but also on the optical networking infrastructure needed to connect increasingly large AI systems.
Qualcomm said the combination of customized processing and connectivity is intended to deliver higher performance, greater efficiency and better cost economics for next-generation AI infrastructure.
KEY QUOTES:
“As AI demand accelerates, data center infrastructure will require advances in both computing and connectivity to deliver greater performance with more efficiency,” said Cristiano Amon, President and CEO, Qualcomm Incorporated. “Qualcomm is pleased to work with AWS on customized silicon and connectivity solutions, bringing decades of leadership in advanced processing and power-efficient compute, to deliver breakthrough performance and enable the next generation of AI infrastructure.”
Cristiano Amon, President and CEO of Qualcomm Incorporated
“This collaboration with Qualcomm Technologies builds on a strong foundation of partnership and reflects our shared commitment to pushing the boundaries of what’s possible,” said Prasad Kalyanaraman, Vice President, AWS. “By working together on customized silicon and advanced connectivity, we’re delivering more performant, efficient, and cost-effective infrastructure for our customers.”
Prasad Kalyanaraman, Vice President at AWS
