Qualcomm Data Center Chip Strategy Targets China Market

Qualcomm unveiled its data center chip lineup on Wednesday, becoming the latest chipmaker to enter the AI processor race in an attempt to challenge market leader Nvidia . According to ZeroHedge, CEO Cristiano Amon told Nikkei Asia the company is eyeing the China market for its data center products, including designing chips specifically for Chinese customers that are in compliance with U.S. export controls. The mobile chip giant is revamping the design of data center processors, which is traditionally powered by graphic processing units (GPUs) and high bandwidth memory (HBM) chips.

Key takeaways
Qualcomm unveiled its Dragonfly data center chip lineup, featuring four product lines: AI accelerators, data center CPUs, custom silicons, and connectivity chips.
The company plans to design China-specific AI accelerators that comply with U.S. export controls, leveraging its existing relationships with Chinese smartphone makers and auto companies.
Qualcomm's high bandwidth compute (HBC) architecture uses near-memory compute design, which the company claims delivers six times the bandwidth per watt versus HBM-based solutions.
The company expects data center products to generate $300 million in revenue in the current fiscal year and $5 billion in fiscal year 2027, targeting more than 5% of a $1 trillion addressable market by 2029.

Table of Contents
What is Qualcomm's Dragonfly data center chip strategy?
How Qualcomm plans to serve the China market
High bandwidth compute architecture explained
Revenue targets and market positioning
Competitive landscape and challenges
Regulatory and geopolitical considerations
Frequently Asked Questions

What is Qualcomm's Dragonfly data center chip strategy?

Qualcomm teased Dragonfly, a dedicated brand for AI data center solutions designed to break Nvidia's grip on AI infrastructure, at the Computex trade show in Taipei early in June 2026. The company unveiled more details about Dragonfly at its investor day in New York on Wednesday. Dragonfly encompasses four product lines: AI accelerators, data center CPUs, custom silicons, and connectivity chips. The source context states that Amon said Qualcomm is working to bring all four data center product lines to China, including customized AI accelerators for the Chinese market that will comply with U.S. export controls limiting advanced AI chips sales above certain thresholds.

The data center compute market is dominated by AI racks powered by Nvidia's GPUs and HBM chips that are produced by South Korean companies SK Hynix and Samsung. Both Samsung and SK Hynix are also working on near-memory and on-memory compute as memory capacity becomes the latest AI deployment bottleneck. Qualcomm's entry into this market represents a strategic diversification from its core mobile chip business, which has long been the company's primary revenue driver.

How Qualcomm plans to serve the China market

Amon stated that Qualcomm has a big business in China, and as the company started to diversify, its partnership with China and its China customers also expanded. According to the source context, Amon added that the relationship with Chinese smartphone makers and auto companies is going to be a strength that Qualcomm will bring on the data center side. However, the CEO emphasized that there are very clear guidelines about how products can be shipped to China, and the company has versions of all of its products that comply with those guidelines.

The source context reports that China accounted for 46% of Qualcomm's revenue in 2025, mostly from smartphone chips. Amon said the company is engaged in conversations and is positively optimistic about the reaction it is getting. Bloomberg reported last month that Qualcomm had struck a deal with ByteDance to supply the Chinese tech giant with custom AI data center chips. The deal is structured to fall within existing U.S. export control thresholds, a design choice that signals Qualcomm's intent to capture Chinese AI demand without irritating the White House.

High bandwidth compute architecture explained

Qualcomm's data center processor features a design that differs from AI racks deployed in data centers. Dubbed high bandwidth compute (HBC), Qualcomm said the near-memory compute design will make its data center chips deliver six times the bandwidth per watt versus HBM-based solutions. Amon said the HBC will be different from the processing-in-memory (PIM) architecture other memory chipmakers are developing. According to the source context, Amon stated that this is a very unique technology that allows the development of 3D-stacking of the DRAM alongside logic that is built for the accelerator, adding that HBC significantly increases available memory, reduces bandwidth bottleneck, and improves compute efficiency.

As the global memory chip crunch continues, Amon said the company has secured enough memory for its data center products in fiscal year 2027 and the new HBC technology will also help ease the memory chip shortage. The CEO said this technology is starting to get interest, with memory vendors, small and large ones, now engaging with Qualcomm and wanting to partner with Qualcomm on HBC. In addition to better compute performance, Qualcomm said the HBC architecture also uses less energy and costs less to own. In video messages, Microsoft and Meta CEOs said the two companies' data centers will be early adopters of Qualcomm's data center chips including HBC and CPUs.

Revenue targets and market positioning

Qualcomm said the first HBC chip will ship with its AI250 data center rack in fiscal year 2027. The company told investors Wednesday the new data center products are expected to bring in $300 million in revenue in the current fiscal year, and $5 billion in fiscal year 2027, which starts in October. Qualcomm estimates the total addressable market for data center chips will be more than $1 trillion by 2029 and the company will take a more-than-5% share of that market. At the Wednesday keynote, Amon said it is never too late for Qualcomm to enter the data center chip business because this is a market that moves very fast, and if a company has technology leadership, there is always room for it.

In addition to AI accelerators powered by the new HBC design, Qualcomm also unveiled a CPU designed for data centers and AI inference, announcing the company has won two major hyperscaler deals for custom-designed data center chips that will bring in meaningful revenue by the end of the year. By fiscal year 2029, however, the company expects handsets to account for only a third of revenue with data center products on par with smartphone chips. For the latest quarter, the source context reports that Qualcomm's handset chip business reported a 13% year-on-year revenue drop to $6 billion, which accounted for 57% of its revenue in the January to March quarter. Its Internet of Things business, which includes PC chips, recorded a 9% sales jump on the year to $1.7 billion.

Competitive landscape and challenges

Amon touted a close partnership with the contract chipmaking giant TSMC that will give Qualcomm a leg up in the data center chip race. According to the source context, Amon stated that as soon as TSMC finishes the mask, Qualcomm goes to production, and goes to production at scale, describing this as the maturity of the company's manufacturing capabilities. It is yet to be seen if Qualcomm can convince investors and customers alike to be a competitive alternative to Nvidia products. Vivek Arya, analyst at Bank of America, said in a note Tuesday that current data center revenues remain de minimis and reliant on Qualcomm proving they are able to bring strong CPU and NPU performance from consumer devices to more complex data center workloads, adding that Qualcomm is entering a fast-growing but hyper competitive AI market full of large incumbents.

Ahead of the investor meeting Wednesday, Qualcomm announced the acquisition of chip software startup Modular Inc. in an all-stock deal valued at nearly $4 billion that will help Qualcomm compete with Nvidia's CUDA ecosystem. In addition to its advanced AI processors, Nvidia's lead in the data center market is solidified by the CUDA computing platform that makes its AI chips more efficient and easier to program. On Wednesday, Nvidia CEO Jensen Huang told the annual shareholders meeting that while his company's systems may not be the cheapest to produce or purchase, Nvidia generates the lowest cost tokens, the highest token throughput, and the most revenues. For readers following broader market updates , Qualcomm's data center strategy represents a significant diversification effort by a major mobile chip supplier.

Regulatory and geopolitical considerations

The U.S. chip giant will likely face similar regulatory scrutiny as Nvidia and others over China exports. The Trump administration unveiled new guidelines for the export of powerful AI chips to Chinese entities outside of China in June 2026. The U.S. Department of Commerce said it will implement license requirements for Chinese companies headquartered in China, even if they are physically located outside the country. Qualcomm's CEO was part of a high-profile business delegation that accompanied Trump when he visited President Xi Jinping in China in May 2026. Amon said Wednesday the company's presence at the leaders summit is an example of what a win-win relationship between the two countries looks like.

Amon said the epicenter of AI agent development is in China, with new agentic use cases emerging across platforms from smartphones and glasses to cars. According to the source context, Amon stated that when he talks about China, he is in a situation right now that he does not know who the mobile customers are anymore, because there are OEMs, but every single AI foundational model company building agents also are customers. Meanwhile, Qualcomm announced a deal with Saudi Arabian AI company Humain, which has committed to deploying 200 megawatts of Qualcomm accelerator racks, beginning this year. The world's leading mobile chip developer, Qualcomm, has long been a leader in premium chips for flagship smartphones such as Samsung Electronics and Xiaomi, and has gradually expanded into the PC market. It announced its first chip for budget PCs at Computex in Taipei earlier in June 2026.

Frequently Asked Questions

What is Qualcomm's Dragonfly data center chip lineup?

Dragonfly is Qualcomm's dedicated brand for AI data center solutions, encompassing four product lines: AI accelerators, data center CPUs, custom silicons, and connectivity chips. The company unveiled the lineup at its investor day in New York on June 25, 2026.

How does Qualcomm's high bandwidth compute architecture work?

High bandwidth compute (HBC) is a near-memory compute design that uses 3D-stacking of DRAM alongside logic built for the accelerator. Qualcomm claims HBC delivers six times the bandwidth per watt versus HBM-based solutions, increases available memory, reduces bandwidth bottleneck, and improves compute efficiency.

What are Qualcomm's revenue targets for data center chips?

Qualcomm expects data center products to generate $300 million in revenue in the current fiscal year and $5 billion in fiscal year 2027, which starts in October 2026. The company estimates the total addressable market for data center chips will be more than $1 trillion by 2029 and targets more than 5% of that market.

How will Qualcomm serve the China market while complying with U.S. export controls?

Qualcomm plans to design China-specific AI accelerators that comply with U.S. export controls limiting advanced AI chips sales above certain thresholds. The company has versions of all of its products that comply with U.S. guidelines and is engaged in conversations with Chinese customers.

What challenges does Qualcomm face in competing with Nvidia?

Qualcomm must prove it can bring strong CPU and NPU performance from consumer devices to more complex data center workloads in a hyper competitive AI market full of large incumbents. The company acquired chip software startup Modular Inc. for nearly $4 billion to help compete with Nvidia's CUDA ecosystem.

What is the significance of Qualcomm's partnership with TSMC?

Qualcomm's close partnership with TSMC allows the company to go to production at scale as soon as TSMC finishes the mask, which CEO Cristiano Amon described as the maturity of Qualcomm's manufacturing capabilities. This partnership is positioned as a competitive advantage in the data center chip race.

Read original source