To The Moon

The AI investment story is changing.

For the past two years, investors have focused heavily on GPUs, data centres and large language models. Companies supplying the computing power behind AI have attracted enormous attention.

But the next stage could look very different.

AI is increasingly moving beyond computer screens and into physical environments. Satellites could process information in orbit, while intelligent robots could use specialised chips to navigate and interact with the real world.

Two upcoming Hong Kong IPOs offer an interesting opportunity to examine these developments:

• Guoxing Aerospace (06935): Bringing AI computing capabilities into space.

• Amicro Technology (02065): Developing the chips and intelligent systems that power mobile robots.

What makes these companies interesting is not simply that they operate in fashionable industries. It is that they represent two different potential destinations for future AI investment.

1. Guoxing Aerospace: What if the next data centre is in space?

Traditional AI infrastructure depends on enormous ground-based data centres, which require electricity, cooling systems, networking and physical infrastructure.

Guoxing Aerospace is exploring a different approach: putting computing capabilities into orbit.

Its business involves AI computing satellites, satellite manufacturing and related space-based solutions.

The potential opportunity is considerable.

As satellite constellations expand, the amount of information collected from Earth observation, communications and other applications will increase. Processing some of this information closer to where it is generated could potentially reduce transmission requirements and enable new services.

However, space computing is still an emerging market.

Satellite launches, maintenance, reliability and the economics of operating computing infrastructure in orbit remain significant challenges.

Guoxing has reported an AI satellite order book worth approximately RMB3.48 billion, covering 114 satellites awaiting delivery as of October 20, 2026.

That provides an indication of commercial interest, but orders are not the same as completed deliveries or collected cash.

The key question is whether the company can turn its technological achievements into recurring, profitable services.

2. Amicro Technology: The intelligence behind the machines

While Guoxing looks towards space, Amicro is focused on bringing intelligence into physical devices.

Its technology includes dedicated AI chips, navigation systems, perception technology and software solutions for intelligent mobile robots.

Think about robotic vacuum cleaners.

Earlier generations mainly followed programmed routes and basic navigation instructions. Newer devices need to recognise obstacles, understand their surroundings and make decisions in real time.

These capabilities require specialised hardware and software working together.

This is where companies such as Amicro fit into the robotics supply chain.

The opportunity could extend beyond household cleaning devices into service robots, commercial applications and eventually more advanced forms of embodied intelligence.

The important distinction is that Amicro does not need to manufacture the final robot to benefit from industry expansion.

It can potentially supply the technology that multiple robot manufacturers need.

However, competition is intense, and customers may have considerable bargaining power. The company must demonstrate that its technology offers enough differentiation to maintain pricing and margins.

3. Two different AI investment cycles

I see an interesting contrast between these businesses.

Space AI depends on the expansion of satellite infrastructure and the development of new commercial applications.

Robotics depends on falling hardware costs, better perception technology and wider adoption of intelligent machines.

Their growth drivers are therefore different.

Guoxing could benefit from increased investment in commercial aerospace and satellite infrastructure.

Amicro could benefit from the growing demand for intelligent devices and automation.

Neither opportunity requires AI spending to disappear from traditional data centres. Instead, they could represent additional markets as AI technology becomes more widely deployed.

4. The valuation question: Is the future already priced in?

This is where IPO investors need to be particularly careful.

Guoxing’s proposed market capitalisation is approximately HK$15–20 billion, while Amicro’s is approximately HK$4–5 billion.

Their businesses are at different stages, so comparing market capitalisation alone does not establish which IPO is cheaper.

Both companies are loss-making.

For Guoxing, investors need to understand how much capital is required to develop and deploy its satellite infrastructure and how quickly those investments can generate recurring revenue.

For Amicro, the focus should be on whether its rapid revenue expansion can eventually produce operating leverage and sustainable profitability.

I would also pay attention to cash flow rather than relying solely on reported revenue growth.

A company can have an impressive order book or rapidly growing sales while still consuming substantial amounts of capital.

5. What could make these IPOs successful after listing?

For Guoxing, I would monitor:

• Actual satellite deliveries against its reported order book.

• Growth in commercial AI computing services.

• Repeat customers and recurring revenue.

• Capital expenditure and cash consumption.

For Amicro, I would monitor:

• Growth in dedicated AI chip sales.

• Expansion into new robotics applications.

• Customer diversification.

• Gross-margin improvement and progress towards profitability.

These are more meaningful long-term indicators than first-day share-price movements.

6. My investment perspective: Follow the commercialisation, not the excitement

I find both business models interesting, but I would approach them differently.

Guoxing is a higher-concept investment. Its potential market is exciting, but the commercial economics of space-based computing still need to be demonstrated.

Amicro has a closer connection to existing commercial products. Its technology serves an established robotics market, giving investors a more tangible way to evaluate demand.

However, existing demand does not automatically mean attractive profitability, particularly when development costs and competition remain high.

I would therefore avoid subscribing simply because one company uses the words Space AI and the other uses robotics.

The final offer price, institutional demand and the amount of capital required to support future growth matter much more.

My conclusion: The next AI winners may come from outside the traditional semiconductor trade.

The first phase of the AI boom rewarded companies supplying computing power.

The next phase could create opportunities in industries that apply that computing power in new environments, from satellites orbiting Earth to robots operating in homes, factories and commercial spaces.

Guoxing Aerospace and Amicro Technology represent two very different ways to participate in this transition.

But the investment opportunity will ultimately depend on commercial adoption, financial discipline and the ability to turn technological innovation into sustainable earnings.

I am interested in both industries, but I would rather see proof of commercial success than pay a premium for a promising story.

For discussion purposes only. Not financial advice.