Alibaba Shifts to In-House Chips Amid U.S. Export Restrictions

Alibaba has started using its own in-house chips in AI systems, reducing reliance on Nvidia components impacted by US export restrictions. This shift is a response to challenging business conditions for Chinese tech companies. Alibaba's financial health is strong, with robust revenue growth and profitability metrics, but there are warning signs such as declining operating margins and a low dividend yield. The company's diverse business model and strong market presence provide a competitive edge, but it must adapt to the tech sector's rapid evolution.

Chinese tech giant Alibaba Group has begun utilizing its own in-house chips for artificial intelligence (AI) training, marking a significant shift away from Nvidia components, which have been impacted by U.S. export restrictions Alibaba and Baidu shift to homegrown chips for AI training - The Information^[1]. This strategic move is part of Alibaba's broader efforts to reduce its reliance on Western semiconductor technology and align with the Chinese government's push for technology self-sufficiency.

Alibaba has started using its Zhenwu processing unit for training smaller AI models since early 2025, while Baidu is testing new versions of its Ernie AI model using its self-developed Kunlun P800 chip Alibaba and Baidu shift to homegrown chips for AI training - The Information^[1]. Despite this transition, both companies continue to use Nvidia’s more powerful and stable chips for their most advanced models, indicating a cautious approach to fully replacing Western technology.

The financial health of Alibaba remains robust, with strong revenue growth and profitability metrics. The company's diverse business model and strong market presence provide a competitive edge. However, there are warning signs such as declining operating margins and a low dividend yield. Alibaba's recent financial moves, including the issuance of $3.2 billion in zero-coupon convertible senior notes maturing in 2032, highlight its strategic focus on fortifying its cloud infrastructure and international business expansion Supplementary ammunition! Alibaba issues $3.2 billion in zero-coupon convertible bonds, with 80% allocated to enhancing cloud infrastructure.^[2]. This large-scale financing initiative underscores Alibaba's commitment to maintaining its leadership in the cloud and AI sectors.

The company's entry into the local lifestyle services market through Gaode Map, which includes over RMB 1.1 billion in subsidies, demonstrates Alibaba's aggressive approach to capturing offline traffic and competing with Meituan and Douyin Supplementary ammunition! Alibaba issues $3.2 billion in zero-coupon convertible bonds, with 80% allocated to enhancing cloud infrastructure.^[2]. This move is part of Alibaba's grand strategy of 'dual-line operations,' where it simultaneously invests in long-term technological barriers through financial instruments like bond issuance and engages in competitive market battles.

Alibaba's latest AI chip unveiling, the Qwen-3-Max-Preview, features over 1 trillion parameters and outperforms previous models, positioning the company as a leading provider of generative AI in China Alibaba Hong Kong Shares Rise As 1-Trillion-Parameter Qwen-3-Max AI Model Debuts—To Challenge OpenAI, Google^[3]. This innovation aligns with Alibaba's strategic push into the generative AI market and its efforts to reduce reliance on U.S. semiconductor technology.

Despite the challenges posed by US export restrictions and regulatory changes, Alibaba's strong fundamentals and diverse business model position it well to adapt to the rapid evolution of the tech sector. The company's strategic initiatives in AI, cloud computing, and international expansion indicate a proactive approach to maintaining its competitive edge.