Qualcomm is in discussions to acquire Tenstorrent, a developer of RISC-V-based AI accelerators and CPUs [1, 2, 3].
The potential acquisition represents a strategic pivot for the U.S. semiconductor company as it seeks to break Nvidia's hold on AI inference workloads. By integrating Tenstorrent's specialized architecture, Qualcomm aims to strengthen its presence in the data-center market, a sector where it has historically trailed behind dominant GPU providers [2, 3, 5].
Reports indicate that the deal would value Tenstorrent, founded by industry veteran Jim Keller, between $8 billion and $10 billion [1, 2]. Some reports specify the valuation could reach up to $10 billion [4].
Tenstorrent focuses on RISC-V, an open-standard instruction set architecture. This approach allows for greater customization of hardware compared to proprietary systems, providing a potential pathway for Qualcomm to diversify its chip capabilities beyond its traditional mobile strengths [3, 5].
Market reaction to the news was immediate. Qualcomm stock jumped over four percent [6] following the reports of the potential takeover.
While the companies have not officially confirmed the deal, the move signals an escalating arms race in AI hardware. Qualcomm's interest in Tenstorrent suggests a desire to move into the infrastructure that powers large-scale AI models, rather than relying solely on on-device AI for smartphones and laptops [3, 5].
“Qualcomm is in discussions to acquire Tenstorrent, a developer of RISC-V-based AI accelerators and CPUs.”
This acquisition would signal a shift in the AI hardware landscape by pairing Qualcomm's massive scale with Tenstorrent's RISC-V innovation. If completed, it allows Qualcomm to challenge Nvidia not just in edge computing, but in the high-margin data-center environment, potentially lowering the barrier for companies seeking alternatives to proprietary AI chip architectures.
