When it comes to artificial intelligence (AI) chip stocks, my guess is that names such as Nvidia, Broadcom, or Advanced Micro Devices come to mind first. While all three companies are playing an integral role supplying data centers with the latest AI hardware, these chip designers have to give a lot of credit to the foundry services of Taiwan Semiconductor Manufacturing (TSM -1.19%).

One thing that makes Taiwan Semi so unique is that its services aren't in demand just from traditional semiconductor companies. In fact, OpenAI just tapped Taiwan Semi for an important collaboration. Let's explore how OpenAI and TSMC are working together and assess why this partnership could be lucrative for Taiwan Semi in the long run.

How are OpenAI and Taiwan Semi working together?

OpenAI's flagship product is a large language model (LLM) called ChatGPT, which uses generative AI to answer queries of all kinds at fast speeds. While ChatGPT's seemingly infinite knowledge gives it the appearance of a modern-day online encyclopedia, there are actually quite a few technical processes going on in the background.

Inside data centers sit advanced chipsets known as graphics processing units (GPU), which run algorithms to process data workloads plugged into ChatGPT in order to feed users answers to their questions promptly. Right now, Nvidia and AMD are the primary players in the data center GPU market -- hence, both companies have enormous leverage over their customers when it comes to pricing power.

However, more companies are now turning to custom silicon solutions in an effort to move away from an overreliance on Nvidia in particular. OpenAI is the latest to explore the development of its own custom chipware -- and (shocker!), it's turning to Taiwan Semi for help.

Image source: Getty Images.

How does this make strategic sense for OpenAI and TSMC?

I don't find OpenAI's choice to work with TSMC a surprise at all. In addition to legacy semiconductor companies, Taiwan Semi has been reported to be working closely with Amazon, Alphabet, Apple, Meta Platforms, and Microsoft, as each of these customers explore alternative solutions to Nvidia architectures.

Regarding OpenAI, remember that the company has close ties to Microsoft in particular -- given the hyperscaler has a $10 billion investment at stake in the ChatGPT maker. To me, OpenAI's interest in its own chipsets is a mere extension of Microsoft's development of custom chipware and its planned AI infrastructure spend of $80 billion this year.

What does this deal mean for Taiwan Semi?

Industry research suggests that Taiwan Semi commands an estimated 60% share of the foundry market -- well ahead of rivals such as Intel and Samsung. In my eyes, TSMC has been able to build its lead primarily thanks to soaring demand for equipment sold by Nvidia, AMD, Broadcom, and other traditional semiconductor businesses.

Now that hyperscalers and adjacent AI developers are also carving out their own pockets of the silicon realm, I think Taiwan Semi is in a lucrative position to kick into a new gear and surge even further ahead of the competition. Another unique angle to its relationships with cloud infrastructure specialists and developers such as OpenAI is that it diversifies TSMC's revenue base.

I think OpenAI is making the obvious choice by choosing to partner with TSMC to bring its chip designs to life. Furthermore, given AI workloads are expected to continue soaring and become more complex over time, my thinking is that additional chip designs will follow in the years ahead.

For this reason, I think the deal with OpenAI could very well lead to even more chipware architectures down the road -- which bodes well for TSMC's long-run prospects.