- Qualcomm is building servers and CPU shelves, looking at AI inference groups in the Hipperscala data center
- The promised CPU of Qualcomm could arrive in 2028, long after the rivals have climbed the infrastructure
- But investors were not excited about the plans of the Qualcomm Data Center, since the prices of the shares fell in the launch of the profits
Qualcomm can finally be ready to go beyond its mobile strength and take a real step in the data centers market.
During its recent earning calls of the third quarter, the company confirmed that it is in “advanced discussions” with a hyperscala customer with respect to the new silicone aimed at server infrastructure.
This could mark a significant change for the company, which for a long time has caused ambitions to carry processors compatible with the arm to cloud computing spaces, but never followed on a scale.
Ambitions grow, but execution will be critical
The CEO of Qualcomm, Cristiano Amon, framed the moment as an opportunity created by the evolution of the workloads of AI, arguing that the cloud service providers focus more and more on efficiency metrics such as tokens per watt and tokens per dollar, not just the yield.
This, together with a broader industry that moves away from the CPU X86 to custom -based personalized alternatives, could finally give Qualcomm the entrance point he has been looking for.
However, the market is already full of established players who have been building infrastructure of the dedicated for years, and Qualcomm will need more than ambition to gain significant traction.
Amon explained that the company is developing “a general purpose CPU” directly addressed to hyperscalers.
“While we are in the early stages of this expansion, we are committed to multiple potential clients and we are currently in advanced discussions with a leading hyperscalero,” he said, “if he succeeds, we hope that the income will begin in the fiscal period 2028”.
Amon also mentioned other projects, including accelerator cards and even complete server racks, additions that are destined to admit AI inference groups, noting that Qualcomm aims to become more than a CPU provider.
Even so, there are questions about the timeline and the competitiveness of such efforts.
While Qualcomm expects the income of this movement to begin around fiscal year 2028, that delay could leave it behind more entrenched rivals as Broadcom, whose personalized accelerator business seems to be extending aggressively.
Despite the optimistic tone of the ad, the reaction of investors was silenced.
Qualcomm actions submerged briefly after the launch of the profits, which suggests that the market remains cautious about the viability of its expansion plans.
The company’s PC business, although it shows modest profits, remains small: its part of Windows premium laptops has reached only nine percent, and there are few indications that it has the fastest CPU in any important computing category.
The competition is also coming in the central segment of Qualcomm, and its rival, Samsung, has indicated that it is preparing to carry its own societies advanced to flagship devices for 2026.
This rivalry suggests that Qualcomm could be trying to diversify out of necessity, instead of a position of force.
For now, the Qualcomm jump to the workstation and space of the data centers remains mostly theoretical.
It remains if you can offer a credible and high -performance CPU and become a serious player in AI infrastructure.
Through the registration
