The rise of AI and what shares to buy next (Dell, HP, Lenovo, etc.)

On the 30th May I tipped various companies including HP, Dell, etc.

I put some limit buys on Trading 212 and managed to pick up HPE for $43. On June 1st it shot up and is now at $59.

Hewlett Packard Enterprise reports record second-quarter results and accelerates its long-term financial goals by two years, underscored by solid demand for artificial intelligence data centers, while Google-owner Alphabet unveils a plan to raise $80 billion in equity capital.

I also had an order for Cisco but it has not yet reached my buy price.

What to buy next?

I am a 'picks-and-shovels' type of investor. I don't want to gamble on new startups or unproven technology.

AI learning is incredibly expensive, you need power-hungry, fast compute power, massive data centers and lots of electricity.

In order to get new customers hooked on AI, it had to be free or cheap. Now most businesses have seen the power of AI now. They have reduced their workforce and taken out large subscriptions to AI providers.

However, they have discovered that it is getting more and more expensive. Each member of staff can easily burn through $100/day of credits if they are not careful. The company is at the mercy of their AI provider. Their data is not private. Any problem with their credit and their AI provider will prevent their staff from working (as they made a lot of their experienced staff redundant).

So most businesses will be looking for an in-house solution where they can control their own costs and their data remains private. They don't need massive AI data centers if they can buy a 90% trained AI model off the shelf and finish the training themselves.

So starting with large companies (and then drifting down to medium and finally small companies), they will buy their own AI servers and infrastructure.

Not only speed, but energy costs will be a major factor. The cost of compute/minute is critical, not just the capital cost.

So we need to look at server companies (Dell, HP, Super Micro, Lenovo, Cisco, Fujitsu, Inspur, etc.) and component companies (esp. lower power CPUs/GPUs that can be used in AI servers). This means looking at ARM, Intel, AMD, nVidia, etc. too.

Lenovo has shot up over 70% this week. They are a dark horse as people don't think of them for servers but they have a strong server division.

Memory and disks will still be needed for smaller servers.

Companies will also need infrastructure and cooling too.

Let's look at my new watchlist app to see who is topping the charts this week...

The trick is to buy undervalued companies. 

Companies Most Leveraged to AI Infrastructure Growth

The companies that stand out are:

1. Vertiv
2. Eaton
3. Schneider Electric
4. ABB
5. Siemens Energy
6. Cummins
7. Johnson Controls
8. Arista Networks

Of those, the most direct beneficiaries of an expansion of private, sovereign and corporate AI infrastructure across Europe, the UK, Japan and the Middle East are probably Vertiv, Eaton, Schneider, ABB and Arista because they sell equipment that is required regardless of whose AI model wins. Whether the customer buys servers from Dell, HPE, Lenovo or a domestic European supplier, they still need power, cooling, racks and networking. That makes these suppliers closer to the "selling shovels during a gold rush" part of the AI value chain.

I already have Vertiv and Siemens which are probably fairly valued at the moment...

ChatGPT has ranked these (not sure I agree though)...

I will investigate the top three over the next few days, but Eaton looks undervalued at first glance and I have put in a buy offer at €398. I missed Lenovo, but also have a buy offer on that now at €2.8.

This is not investment advice - please do your own research!