Nvidia’s goldmine needs shovel-makers. Here’s one place to find them

There are no Asian equivalents of the AI-driven ‘Magnificent Seven’ US tech stocks. But there are AI opportunities in Taiwan if you know what to look for. J O Hambro’s SAMIR MEHTA explains

• AI hardware makers have big technical challenges to overcome
• Taiwanese companies such as Jentech have the knowhow
• Find out about Samir Mehta’s Pendal Asian Share Fund

I WAS amused recently to read about the viral resurgence of the iconic Stanley Thermos from a dusty attic relic to an Instagram darling.

Similarly, an ordinary home cleaning product became a coveted possession almost overnight in the UK thanks to a Tik Tok “cleanfluencer”.

Scalability or virality is one of the most desired attributes for a business, especially when accompanied by minimal asset intensity.

The powerful “network effect” of social media propels this concept in ways that boggle the mind.

These are not software products or video games, which have an innate quality of achieving scale if conditions are right.

Making physical stuff is hard to do at scale while ensuring quality and consistency, managing supply chains and such.

Yet we can see a similar effect today among some Asian companies benefiting from the AI story.

Asia does not have a ‘Magnificent Seven’ tech stock equivalent.

But a select group of Taiwanese hardware manufacturers can stake claims as the unsung heroes of an AI-driven shift.

Semiconductor maker TSMC is well known as a critical supplier in this rapidly growing market.

Today I want to outline another example – Taiwanese component maker Jentech Precision Industrial. 

But first, an update on how the rapidly changing AI landscape is impacting the manufacturing of chips and servers.

And who else to help us with this question but Google’s AI app Gemini?

I asked Gemini to list the key challenges for manufacturers of AI chips and AI servers as these components become the workhorses of AI.

These are the challenges Gemini highlighted for AI chip manufacturers:

1. Complexity and Cost:

AI chips often employ specialised architectures like “tensor” cores, which are more intricate to design and fabricate compared to standard computer chip cores. This means higher manufacturing costs and lower yields (usable chips per wafer).

Pushing performance boundaries requires ever-smaller transistors, requiring expensive and cutting-edge lithography techniques that aren’t readily available at scale.

Fabrication complexity and cost is also increased by the need to integrate high-bandwidth memory directly onto a chip for improved communication.

2. Heat dissipation:

AI computations demand significant power, leading to intense heat generation. Traditional packaging and cooling solutions often don’t suffice, requiring innovative thermal management methods.

And here are the challenges Gemini nominated for AI server manufacturers:

1. Scalability and efficiency:

Combining different chips (CPUs, GPUs, specialised AI accelerators) within a server demands new designs for efficient communication and data flow.

Find out about

Pendal Asian Share Fund

Read more

Power and cooling is again a challenge. Accommodating high-power AI components necessitates robust power delivery and intricate cooling systems, impacting server design and layout.

Server hardware also needs to be tightly coupled with optimised AI software for efficient resource use and minimised bottlenecks.

2. Changes and new designs:

There is a trend away from rigid, fixed setups to modular designs that allow for flexible configurations based on specific AI workloads. This improves efficiency and scalability but requires innovation.

Traditional air cooling might not be enough for high-density AI servers. Liquid cooling systems, while effective, pose challenges in terms of cost, complexity and potential leaks.

New high-bandwidth, low-latency interconnect technologies are needed to facilitate efficient communication between diverse processing elements inside a server.

What this means for Asian tech leaders

I apologise for quoting at length but I think it’s important to understand the complexity of the challenges posed by the fast-growing AI industry.

Gemini’s list above tells us there is a great deal of development still to come if we are to reap the benefits of AI in ways that are meaningful and significant.

I believe Asian economies are likely to benefit from this process in ways that are not yet recognised by international investors.

We own several companies across this spectrum of innovation and change in manufacturing.

One of them is Jentech Precision Industrial.

With poor analyst coverage, Jentech was lumped in with machining and casting businesses.

It was seen as another component manufacturer in a long list of components in a very cyclical industry.

But changes in AI server technology have led to significant added value and potentially an expanded business cycle.

The simplest way to understand this is the likelihood of a PC and device replacement cycle (in companies and consumers) driving volumes.

Also, a monstrous increase in “active server pages” required for AI-related critical components.

It’s difficult to cite averages, but I believe there could be an almost 8-to-10 times increase in realisations for some component makers compared to normal server components.

Jentech crafts intricate “heat spreaders” which keep AI servers cool under pressure.

Its products help dissipate heat from integrated circuits and strengthen thermal modules while avoiding warping.

Companies like Jentech are defined by years of manufacturing know-how.

Another secret to competitiveness is longer durability of their moulds which reduces costs over the long run.

As Jentech’s client AMD moved to larger heat spreaders, Intel had no choice but to approach Jentech too.

Now Jentech’s thermal solutions play an important role in data centres powering AI breakthroughs and even find uses in self-driving cars.

Shovel-makers for Nvidia’s goldmine

Think of it this way: While Nvidia stole the spotlight, Taiwanese companies like Jentech were busy building the props and the stage machinery, ensuring the show went on without a hitch.

Manufacturers like Jentech are the shovel-makers for Nvidia’s goldmine.

From what I understand, most of Nvidia’s clients want to ensure they are not solely dependent on Nvidia – which means the rest of the Magnificent Seven could become potential customers for these Taiwanese companies in time.

Nothing in technology can be taken for granted.

But many of these Taiwanese firms are the go-to partners of choice with very few alternatives when it comes to leading-edge technologies.

About Samir Mehta and Pendal Asian Share Fund

Samir manages Pendal Asian Share Fund, an actively managed portfolio of Asian shares excluding Japan and Australia. Samir is a senior fund manager at UK-based J O Hambro, which is part of Pendal Group.

Pendal Asian Share Fund aims to provide a return (before fees, costs and taxes) that exceeds the MSCI AC Asia ex Japan (Standard) Index (Net Dividends) in AUD over the medium-to-long term.

Pendal is a global investment management business focused on delivering superior investment returns for our clients through active management. 

Contact a Pendal key account manager here