퀄컴과 AWS는 최대 1.6T 처리량을 목표로 하는 광 연결 솔루션을 포함한 대규모 AI 데이터센터용 맞춤형 실리콘 개발을 위해 파트너십을 체결했다.
Qualcomm Technologies and Amazon have announced a multi-generational partnership to develop customized silicon for large-scale AI data centers, with a specific focus on AI inference. The collaboration will also encompass optical connectivity solutions capable of reaching speeds up to 1.6T. By combining Amazon’s AI infrastructure expertise with Qualcomm’s semiconductor design capabilities, the initiative aims to deliver highly optimized hardware for next-generation computing workloads.
Development of the new optical connectivity solutions will leverage Qualcomm’s SerDes and optical digital signal processing (DSP) technologies. In parallel, Qualcomm plans to expand its internal reliance on AWS AI infrastructure, integrating services such as the Amazon Bedrock platform to accelerate its electronic design automation (EDA) workloads.
“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.”
“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,” added 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.”
The partnership aligns with AWS’s ongoing expansion of its custom silicon portfolio, which includes the Trainium and Inferentia chips engineered specifically for AI workloads, alongside its Graviton line of CPUs. In December 2025, AWS introduced the Trainium3 generation, marking the company’s first 3nm AI chip. According to AWS, the Trainium3 UltraServer delivers up to 4.4 times more compute performance, four times greater energy efficiency, and nearly four times the memory bandwidth compared to its predecessor.
On the silicon front, Qualcomm recently expanded its data center portfolio in June, unveiling the Dragonfly C1000 CPU, High Bandwidth Compute (HBC), and the Dragonfly AI300 inference accelerator. Meta has already committed as a customer for the Dragonfly C1000 CPU, planning to deploy the processors across its next-generation server fleet.