Tired of ads? Enjoy an ad-free experience by signing up.
  • Insights
    This article was written by a TIA community member. Insights pieces undergo the same rigorous editorial process that newsroom-produced articles have.
Jacob Cooke · · 4 min read

Alibaba’s AI capex spike isn’t the story investors think

The market hasn’t been kind to Alibaba Group since the Chinese firm released its earnings for the June quarter on August 20. After opening at around US$130, the stock tumbled that trading day by 6%. As of September 8, it is trading down at around US$113 on the New York Stock Exchange.

A quick glance at the headline numbers from the earnings report – profit down by 76%, capital expenditures up by 75% – explains the market reaction.

But I have a different view: the figures show that Alibaba is positioned to capture an enormous share of China’s AI demand over the next few years.

Digital collage by Ulla, photos courtesy of Shutterstock

In fact, if you dig into the numbers, Alibaba’s approach to AI might already be paying off.

Where the full-stack bet is landing

It’s important to point out that Alibaba is making massive investments across the entire AI stack: proprietary silicon, cloud infrastructure, frontier models, consumer applications, and enterprise applications.

In other words, Alibaba owns every layer of the AI supply chain, from the chips powering AI and the cloud renting out compute to the models themselves and the apps that people use.

This setup is different from most Western AI companies that occupy distinct layers of the AI stack. Nvidia builds the chips, frontier labs like OpenAI develop the models, and hyperscalers like Microsoft provide computing power.

See also: A quiet threat is closing in on China’s open-source AI

Alibaba is one of the few global companies with a leading position across the entire stack, which means it stands to benefit from growing demand across every layer.

What does the company’s full stack look like in practice?

T-Head, its in-house semiconductor arm, designs its own AI processors. Zhenwu, the current generation of AI chips from T-Head, goes into Alibaba’s data centers, which get rented out to enterprises through Alibaba Cloud. Finally, Qwen compares well against the top frontier models from China and the US.

As such, growing customer demand at any layer will translate into revenue growth for Alibaba. The latest earnings report shows that this strategy is working.

Where it all connects

Stay ahead in Asia’s tech landscape

This is premium content. Subscribe to read the full story.

Why subscribe?

The market punished Alibaba’s earnings. But a closer look at the firm’s chips, cloud, and AI model suggests the reaction got the story backward.

📖 For learners / 👍 Starter

Lite

US$4.92/month

Billed annually at US$59/year

Get instant access to this article and more every month

4

4 premium content

Unlimited news briefs & articles

10

10 company database access

Ad-free reading experience

Just US$0.17 per day

Cancel anytime

🧠 For professionals / ⭐ Best value

CoreBest value

US$16.58US$14.92/month

Billed annually at US$179.10 on the first year

Get instant access to this article and more every month

Unlimited premium content

Unlimited news briefs & articles

Unlimited company database access

Ad-free reading experience

Just US$0.55 per day

Save US$19.90 on the first year. Cancel anytime

Our subscriber community includes professionals from these companies:

Stay updated on the go with our mobile app.

Get latest insights with smoother, more personalized experience through TIA mobile app.

Community Writer

Jacob Cooke

Jacob Cooke is co-founder and CEO of WPIC Marketing + Technologies.