Qualcomm (QCOM, Financial) has announced plans to launch processors specifically designed for data centers to power artificial intelligence (AI), aiming to integrate with NVIDIA's GPUs. NVIDIA's GPUs have become essential for data centers, which are crucial for training large AI models that support applications like chatbots. Typically, GPUs work alongside CPUs, a market currently dominated by Intel and AMD.
Qualcomm's new custom data center CPUs will connect with NVIDIA's GPUs and software, a critical move for any player entering the data center market due to NVIDIA's significant role in AI. This announcement marks Qualcomm's renewed effort in the data center space, following past challenges. In 2021, Qualcomm acquired Nuvia, a company designing processors based on Arm architecture, which is pivotal for its data center CPU strategy.
The data center CPU market remains highly competitive, with major cloud companies like Amazon and Microsoft designing their custom CPUs, while Intel and AMD hold substantial market shares. Recently, Qualcomm signed a memorandum of understanding with Saudi AI company Humain for data center development, supported by the Saudi Public Investment Fund.
Qualcomm's expansion into data centers is part of its diversification strategy, traditionally reliant on smartphone processors and modems. Under CEO Cristiano Amon, Qualcomm has also ventured into automotive and PC chips, areas traditionally led by Intel. Amon revealed that over 85 PC designs using Snapdragon X series chips are in development, with a new PC chip set to launch in September. Qualcomm positions its chips as energy-efficient, capable of running AI programs on devices rather than the cloud, offering faster response times and enhanced security.
