Integrated Chip with Nvidia: Next Generation Move from Qualcomm

Known for its success in mobile processors, Qualcomm is now making a major comeback to the data center industry this time with a bold new strategy: developing AI focused chips for next generation servers. And the company isn’t going it alone; it’s partnering with none other than Nvidia, one of the biggest names in AI hardware.
Unveiled at Computex 2025, the new chips will support Nvidia’s NVLink-C2C interconnect technology. This means Qualcomm’s processors will be able to directly communicate with Nvidia’s powerful GPUs and Grace CPUs bringing seamless AI performance and energy efficiency to modern data centers.
A Qualcomm spokesperson shared: “We’ve taken the energy efficiency we mastered in mobile and are now applying it to data center-scale AI workloads. This partnership with Nvidia is just the beginning.”
A Second Try, but This Time with a Clear Focus
This isn’t Qualcomm’s first attempt at the server market. Back in 2017, the company tried to break in with its Centriq processors, but the project was shelved due to financial and legal challenges. Things have changed since then.
Qualcomm has brought in top talent especially former Apple silicon engineers and now the focus is sharper than ever: AI acceleration.
The new chips are optimized for tasks like machine learning training, inference, and real time data processing.
Designed with low power operation in mind, they aim to appeal to both large scale cloud providers and enterprises building AI driven infrastructure.
Another Strategic Partner: Ampere
Qualcomm’s collaboration isn’t limited to Nvidia. The company is also teaming up with Ampere Computing to build cost efficient AI servers.
These systems will combine Ampere’s Arm-based CPUs with Qualcomm’s Cloud AI 100 Ultra chips, targeting enterprise users who need AI capabilities without the sky high power demands and hardware costs.
This partnership is all about delivering scalable, energy conscious alternatives for AI inference helping companies adopt artificial intelligence without the need for massive infrastructure upgrades.
Sources:
- Reuters – Qualcomm to build data center chips that connect with Nvidia
- Barron’s – Nvidia, Broadcom drive next-gen AI chip ecosystems
- SiliconANGLE – Ampere and Qualcomm collaborate on cloud AI systems