Qualcomm and AWS Team Up on Custom AI Silicon and 1.6T Optics

Qualcomm and AWS custom AI data center silicon with 1.6T optical connectivity

Qualcomm and Amazon Web Services have announced a multi-generation collaboration to develop customized silicon and high-speed optical connectivity for large-scale AI inference infrastructure.

The Sept. 8 agreement expands Qualcomm’s push into data centers and gives AWS another path to customized compute as inference workloads grow in scale and complexity.

Custom silicon across multiple generations

The companies will work together on customized silicon for AWS AI infrastructure across multiple product generations. Qualcomm says the collaboration will draw on its experience in low-power computing, system design and semiconductors.

The embedded video comes from Qualcomm’s verified YouTube channel and is in English. Qualcomm CFO and COO Akash Palkhiwala discusses the AWS collaboration and the company’s approach to AI data-center infrastructure.

Optical links extend to 1.6T

Compute is only one part of the agreement. Qualcomm and AWS also plan to collaborate on advanced optical connectivity extending to 1.6T and future generations. Qualcomm says the work will use its SerDes and optical-DSP technologies to address growing bandwidth requirements inside AI facilities.

That makes the partnership relevant to the same Ethernet and optical-fiber scaling problem BitcoinVersus.tech has been following as GPU clusters become larger and more distributed.

AI inference pushes efficiency higher on the priority list

Inference systems have to generate useful tokens continuously rather than simply maximize peak benchmark performance. Qualcomm is positioning its data-center roadmap around performance per watt, memory efficiency and system-level integration as electricity becomes a major constraint on AI expansion.

Qualcomm also plans to expand its use of AWS AI infrastructure, including Amazon Bedrock, for electronic-design-automation workloads. The company says the goal is to reduce chip-design cycles.

The announcement complements BitcoinVersus.tech coverage of Marvell’s 2 nm optical networking, NetApp’s planned PEAK:AIO acquisition and Wise and Navitas’ GaN and SiC AI power collaboration.

Sources: Qualcomm and Reuters.


Advertisement:

BitcoinVersus.Tech Editor’s Note & Disclaimer:
We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. BitcoinVersus.tech provides news, technical analysis and educational information and does not provide financial or investment advice. Verify technical specifications and operational requirements with the manufacturer before making purchasing, deployment or investment decisions. If you would like to support to help further secure the integrity of our research initiatives, please donate here: 3C9o19EH5HSiwEPyCTmEKzxhNCbo2X6TTb

8 responses to “Qualcomm and AWS Team Up on Custom AI Silicon and 1.6T Optics”

  1. […] infrastructure surrounding the accelerator, a theme also visible in BitcoinVersus.tech coverage of Qualcomm and AWS custom AI silicon and Axelera’s Europa […]

    Like

  2. […] discussed in BitcoinVersus.tech reporting on Giga Computing’s 700 kW AI factory, Qualcomm and AWS custom AI silicon and onsemi’s denser AI-rack power […]

    Like

  3. […] is evolving just as quickly. 1.6T optical links, 3.2T-class networking technology and massive silicon-photonics manufacturing expansion all point […]

    Like

  4. […] Those constraints echo terrestrial infrastructure topics BitcoinVersus.tech follows in high-bandwidth data-center networking, dense fiber infrastructure, energy-efficient network switches and 1.6T optical networking. […]

    Like

  5. […] interconnected. Alibaba is vertically integrating models, chips and data centers, while custom AI silicon is being paired with hyperscale cloud infrastructure. At the networking layer, 1.6T coherent optics and 6.4T CPO are addressing the bandwidth required […]

    Like

  6. […] recently covered Qualcomm’s work with AWS on custom AI silicon and 1.6T optical connectivity. Dragonfly shows why those pieces belong together. An inference rack is increasingly a coordinated […]

    Like

  7. […] model resembles the direction BitcoinVersus.Tech covered when Qualcomm and AWS paired custom AI silicon with 1.6T optical connectivity. The value is increasingly in the entire compute fabric, not simply the processor […]

    Like

  8. […] technical side is expanding just as quickly. BitcoinVersus.Tech recently covered Qualcomm and AWS collaborating on custom AI silicon and 1.6T optical connectivity. Built Together addresses the other side of that buildout: the people and infrastructure outside […]

    Like

Leave a Reply