Qualcomm Eyes $8B-$10B AI Deal: Can Tenstorrent Acquisition Fuel Next Growth Phase?

For Qualcomm (QCOM), the smartphone market has been a cash cow, but slowing upgrade cycles have forced the chipmaker to seek new growth engines. Now, a reported $8-10 billion acquisition of Tenstorrent could be the catalyst that propels it into the AI data-center race, posing a direct challenge to Nvidia.

Qualcomm, the dominant force in mobile chips, is reportedly in advanced discussions to acquire AI chip startup Tenstorrent for an estimated $8 billion to $10 billion, as it aggressively seeks to pivot beyond the maturing smartphone market. The deal, first reported by The Information, remains non-binding and could still break down, but it underscores the urgency with which Qualcomm is pursuing artificial intelligence as a core pillar of its future growth.

“

Qualcomm, the dominant force in mobile chips, is reportedly in advanced discussions to acquire AI chip startup Tenstorrent for an estimated $8 billion to $10 billion, as it aggressively seeks to pivot beyond the maturing smartphone market.

Tenstorrent is far from a typical AI startup. Headed by legendary chip designer Jim Keller—the architect behind Apple's A-series and Tesla's autonomous driving chips—the company specializes in high-performance AI accelerators for training and inference workloads. Its intellectual property spans data-center-grade hardware, a RISC-V-based architecture, and a licensing model that departs from the proprietary ecosystems of Nvidia. For Qualcomm, which built its empire on Snapdragon processors for smartphones, acquiring Tenstorrent would be a strategic shortcut to relevance in the data-center AI market, where Nvidia controls an estimated 80% of the accelerator market.

The smartphone market that made Qualcomm a $150 billion company is facing structural headwinds: lengthening replacement cycles, incremental hardware improvements, and price pressure from device makers. In response, Qualcomm has already expanded into automotive (Snapdragon Digital Chassis) and data-center CPUs (via its Nuvia acquisition). Tenstorrent would fill the AI gap, complementing Qualcomm’s edge AI capabilities on mobile and automotive with a credible data-center AI story. It would also immediately add Keller’s engineering talent to Qualcomm’s ranks—a major draw given the scarcity of chip design luminaries.

From a competitive standpoint, the acquisition would directly challenge Nvidia’s dominance by offering an open, licensable alternative to CUDA. Tenstorrent’s approach, which emphasizes modularity and customization, could appeal to hyperscalers seeking to reduce dependency on a single supplier. Moreover, Qualcomm’s existing relationships with automakers and its growing autonomous driving chip ambitions dovetail with Tenstorrent’s technology, potentially creating a unified AI platform from the edge to the cloud.

Investors, however, will scrutinize the price. At up to $10 billion, the deal would be Qualcomm’s largest since its attempted $44 billion acquisition of NXP Semiconductors in 2016 (which was blocked). While Qualcomm’s balance sheet is robust, a cash outlay of this size could pressure margins and slow share buybacks. The talks’ tentative nature also introduces uncertainty; similar high-profile chip deals, such as Nvidia’s attempt to acquire Arm, have collapsed under regulatory scrutiny. The Biden-era CHIPS Act and national security considerations around AI technology could complicate approvals.