ALTHOUGH artificial intelligence (AI) is a big theme these days, Malaysian listed technology companies are not known to be benefitting in a big way.
But things could be changing, as larger listed firms are attempting to make forays into AI-related segments.
In addition, a handful of Penang-based smallish tech companies, which have come to the market more recently, have been attracting investor attention due to their AI-related sales growth.
First, the big boys.
Take the case of ViTrox Corp Bhd – it is now embedding some AI elements into its established automated vision inspection systems which it had built over the years, and which caters to the semiconductor and electronics industries.
The AI aspects of their systems, which detect defects better and enable predictive maintenance, are all aimed at servicing the booming AI server and advanced packaging demand.
Pentamaster Corp Bhd is in the early stage of growing its AI-related business.
The group has been a long-time player in providing automated test equipment (ATE) and factory automation solutions, particularly for the automotive sector.
However, following a cooling of the electric vehicle market, Pentamaster has been shifting its focus towards higher-growth areas within the semiconductor value chain.
This includes developing new testing and automation solutions for AI accelerators, advanced chip packaging and data centre (DC) infrastructure.
Pentamaster has described 2026 as its “execution year”, as the group is developing and delivering several prototype machines related to this.
MI Technovation Bhd, another ATE company, is gaining from the rise of high-bandwidth memory and advanced packaging as its products see a slight rise in demand.
But many investors are turning to the newbies on the block, where AI exposure may be more direct.
“Companies such as THMY Holdings Bhd, Northeast Group Bhd and Ambest Group Bhd are appealing primarily because they sit in critical, high-value segments of the semiconductor ecosystem that are benefitting from structural AI-driven demand,” Tradeview Capital Sdn Bhd senior analyst Tan Jia Hui tells StarBiz 7.
THMY came to the market with a bang last October, opening with a 158% premium on its first day of trading. Its shares now trade at RM1.30 apiece, 319% above its IPO price of 31 sen.
The group has a market capitalisation of RM1.15bil, and trades at a historical price-to-earnings ratio (PE) of 27.43 times for the financial year ended March 31, 2025.
An automated test solution provider, THMY is “the most direct AI proxy within the local tech space”, notes Tan.
“This is given its exposure to hyperscaler-driven chips,” Tan says.
THMY designs and develops equipment that is used to test for defects and performance in the electrical and electronics manufacturing process.
The equipment is involved in the final steps of the manufacturing process prior to the product being packaged.
One fund manager says as long as AI continues to scale – requiring more chips, servers and DCs – testing demand will continue to grow.
For instance, global technology companies are continuously rolling out new generations of AI chips.
Each new generation of an AI chip brings different testing requirements as changes in printed circuit board design, power requirements and system architecture mean new testers are needed, and this in turn supports the demand for testing solutions.
“THMY is also closely linked to application-specific integrated circuits (or ASIC, essentially custom-built chips designed to perform one specific task very efficiently), which is where much of the industry is heading.
“Many large technology players are developing their own custom chips – from AI-focused firms to major cloud providers and social media platforms – as they look to reduce reliance on dominant suppliers like Nvidia Corp and thereby lower overall computing costs.
“Hence, ASIC is another avenue of growth for THMY in terms of a new customer base,” he adds.
Meanwhile, a further company in focus is Northeast.
Listed on the ACE Market of Bursa Malaysia in October 2024, Northeast is a maker of precision engineering components – tailored, among others, for the photonics space.
Photonics entails the use of light (photons) through lasers and optical fibres to transmit, process or detect information.
As computing power increases, especially with AI, traditional electrical systems are hitting speed bottlenecks and heat or power issues.
Photonics enables faster data transmission while using less power, making it a critical backbone for next generation computing and AI infrastructure.
In DCs, for instance, photonics is found in optical transceivers or fibre optics to move data between servers.
“The group offers a more indirect but durable AI angle through its strong positioning in photonics, which underpins high-speed data transmission and optical connectivity essential for AI infrastructure, making it a structural beneficiary of rising data traffic and network upgrades,” says Tradeview’s Tan.
Since its listing, Northeast has enjoyed a steady rise in profits, reporting seven quarters of bottom line growth mostly led by photonics-related sales.
As for Ambest, whose shares have more than doubled since its listing just two months ago, the group provides upstream exposure to the semiconductor equipment cycle, supplying mission-critical precision components to original equipment manufacturers.
Hence, the company benefits from AI-led wafer fab and equipment capital expenditure.
According to the fund manager, precision engineering players have seen a strong ramp-up over the last one and a half months, with their capacity fully utilised.
This, he says, is underpinned by DCs that are ordering a lot more chips.
As chip demand rises, orders for front-end semiconductor equipment – supplied by global equipment manufacturers – also increase, and in turn drive demand for precision components used within these machines.
“Anyone within these supply chains is seeing a strong increase in orders,” he says, adding that from a valuation perspective, precision engineering companies such as Northeast and Ambest are trading at relatively reasonable multiples.
Both stocks trade at historical PEs of 31.3 times and 36.4 times, respectively.