Analyst(s): Olivier Blanchard
Publication Date: May 22, 2025
Qualcomm is re-entering the data center market with custom central processing units (CPUs) designed to connect with NVIDIA’s GPU and software infrastructure. The announcement at Computex 2025 highlights the company’s focus on building AI-optimized processors that support NVIDIA’s stack, a critical requirement for data center relevance amid AI proliferation.
What is Covered in this Article:
- Qualcomm re-enters the data center CPU market with custom Arm-based processors
- The new chips will interoperate with NVIDIA GPUs and software stacks
- The initiative is supported by Qualcomm’s 2021 acquisition of Nuvia
- Qualcomm signed an MoU with Saudi-based Humain for AI data center development
- The announcement reflects a larger diversification strategy beyond mobile
The News: Qualcomm is making its way back into the data center CPU market, sharing plans to launch custom-built processors designed to work alongside NVIDIA’s GPUs and complete AI software stack. The idea is to deliver energy-efficient, high-performance AI computing that fits neatly into NVIDIA’s rack-scale infrastructure.
As part of that expansion, the company signed a memorandum of understanding with Saudi-based Humain, opening a path for joint AI data center development. This move targets global infrastructure growth and international presence.
Qualcomm’s Arm-Based Data Center CPUs To Smoothly Integrate With NVIDIA
Analyst Take: Qualcomm’s re-entry into the data center market marks a more calculated move than its earlier, less successful attempt with Centriq. Rather than going it alone, the company is now aligning with NVIDIA’s ecosystem. While few specifics have been shared, this signals a long-term play that builds on Qualcomm’s strengths in mobile and its newly added chip design expertise from Nuvia.
Does Qualcomm have a shot in an already crowded and competitive market? The answer hinges on Qualcomm’s differentiated value proposition: Qualcomm hopes that its processors’ power efficiency (performance per watt) characteristics will catch the market’s attention, specifically given the high energy consumption (and associated costs) of AI-focused data centers. Improving power efficiency is a critical imperative for data center operators, and this very performance advantage is where Qualcomm has traditionally had the upper hand against its competitors. Second, Qualcomm’s on-device AI capabilities, which include low latency and built-in security, have greatly improved in recent years, making the company a legitimate player in the AI chip space.
Will Qualcomm CPUs ultimately be faster or better than incumbent competitors’ top-tier options? That remains to be seen. But if Qualcomm can win on TCO and/or ROI by making data centers more efficient and cost-effective, especially as a natural Arm-based complement to NVIDIA GPUs, this advantage could carve out a significant and disruptive niche in the market for Qualcomm.
A Re-Entry Backed by New Capabilities
In 2018, Qualcomm’s first shot at the data center space fell apart due to weak adoption and internal roadblocks. This time, things are different. We can trace a portion of this comeback to Qualcomm’s 2021 acquisition of Nuvia, a chip design firm based on Arm architecture, an essential element of Qualcomm’s broader diversification strategy under CEO Cristiano Amon. Nuvia’s acquisition helped Qualcomm accelerate work on its next generation of CPU IP, most notably in the PC and Automotive sectors and the data center segment.
Thanks not only to the acquisition of Nuvia but also to the in-house IP developed to complement Nuvia’s solutions, Qualcomm now has a much stronger footing in Arm-based CPU development. The timing also aligns with market forecasts, which predict the data center CPU sector will more than double from $14.16 billion in 2025 to over $30 billion by 2035. Qualcomm appears ready to jump in with better-aligned tools and partnerships.
Qualcomm-Arm Litigation
Qualcomm and Arm have been embroiled in litigation surrounding technology licensing following Qualcomm’s acquisition of Nuvia. (Look for an upcoming deep dive into the status and outcome of the case in the coming weeks.) As things stand, a jury found that Qualcomm had not violated the terms of its technology licensing agreement with Arm when it acquired Nuvia. Arm’s aggressive legal and PR campaign to try to block Qualcomm from using Nuvia-derived IP as well as any related IP that Qualcomm had developed after the acquisition failed, leaving no question as to Qualcomm’s ability to move forward with its CPU roadmap.
Tapping into NVIDIA’s Ecosystem
Instead of positioning itself as a direct rival, Qualcomm’s strategy now focuses on compatibility. Its upcoming chips are being tailored to integrate fully with NVIDIA’s end-to-end systems, from hardware to software and large-scale infrastructure. This approach stands in contrast to the Centriq 2400, which targeted general-purpose server use on its own, and to the Cloud AI 100, which competed head-on with NVIDIA’s inference solutions without ecosystem support. Now, Qualcomm is designing CPUs that work inside NVIDIA’s dominant architecture. That could improve its chances with mid-tier data center operators already committed to NVIDIA’s tech stack.
Expanding Global Relevance Through Strategic Partnerships
The agreement with Humain, a Saudi firm backed by the Public Investment Fund, reflects Qualcomm’s interest in growing its commercial and geopolitical relevance in a rapidly developing tech hub. The partnership will focus on rolling out hybrid AI data centers powered by Snapdragon and Dragonwing processors. This strategy mirrors moves made by other U.S. tech players that have teamed up with Middle Eastern governments to support national AI infrastructure projects, giving Qualcomm a foothold in broader government-led initiatives.
Unfinished Roadmap Leaves Open Questions
Despite the bold pitch, Qualcomm hasn’t yet shared when its new CPUs will hit the market or their specs. The company seems to be betting on long-term growth in AI and cloud infrastructure, although it still faces tough competition from Intel, AMD, and custom chip programs from Amazon and Microsoft. While its partnership with NVIDIA might smooth the path to adoption, Qualcomm will need to show real progress before it can claim significant market share.
What to Watch:
- Qualcomm has not disclosed a CPU roadmap or launch date, leaving timing and execution unclear
- NVIDIA’s dominance in AI infrastructure sets a high technical and ecosystem integration bar
- Qualcomm must compete with established CPU players like Intel and AMD, as well as in-house chips from hyperscalers
- Future diversification efforts could include acquisitions to support scaling on the data center side
Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.
Other insights from Futurum:
Qualcomm’s Q2 FY 2025 Earnings: YoY QCT Revenue Growth Momentum Continues
Qualcomm’s X85 5G Modem Sets New AI-Powered Speed and Efficiency Benchmark
MWC 2025: Qualcomm and IBM Raise GenAI Competitive Stakes
Author Information
Olivier Blanchard is Research Director, Intelligent Devices. He covers edge semiconductors and intelligent AI-capable devices for Futurum. In addition to having co-authored several books about digital transformation and AI with Futurum Group CEO Daniel Newman, Blanchard brings considerable experience demystifying new and emerging technologies, advising clients on how best to future-proof their organizations, and helping maximize the positive impacts of technology disruption while mitigating their potentially negative effects. Follow his extended analysis on X and LinkedIn.
