Cover Story: Resetting Selangor’s digital infrastructure
This article first appeared in Digital Edge, The Edge Malaysia Weekly on January 12, 2026 - January 18, 2026
The Selangor government is executing a major pivot, moving away from its reliance on the premium for state-owned land and low-value manufacturing towards a high-tech, data-driven economy.
Having learnt hard lessons from the recent pandemic, the state government aims to capitalise on its key resources — a sizeable talent pool and stable financial reserves — to move into complex industries, create high-paying jobs and build robust digital infrastructure, while enhancing its ability to provide better public services.
The move is anchored in policies set out under the five-year First Selangor Plan, which concluded in December 2025. Its successor, the Second Selangor Plan, which focuses on implementation and building on the first phase, is expected to be tabled in June. Among the key thrusts of the plan is the establishment of economic clusters aligned with the state’s regional strengths, including aerospace, services and, more recently, semiconductor integrated circuit (IC) design.
While the Selangor government is looking to create smart and liveable cities, its digital infrastructure is among the critical linchpins supporting these complex industries. To address this, the state announced on Oct 6 last year that it was committing RM160 million to the Selangor Digital Future Action Plan 2026-2030 (DxF).
With DxF, the Selangor government aims to resolve legacy issues while pursuing “moonshot” solutions. Running parallel to these initiatives is a dedicated semiconductor-centric policy that seeks to attract global firms and even Malaysian IC engineers working abroad to return home and set up shop in the state. The state has pledged a RM100 million venture capital fund, on top of federally backed incentives, to ensure the industry has a proper runway to grow and mature.
In exclusive interviews with Digital Edge, state executive councillors Fahmi Ngah (Islamic affairs and innovation culture) and Ng Sze Han (investment, trade and mobility) talk about the comprehensive roadmap that will see the state transition from being a land administrator and regulator to being a venture capitalist and digital infrastructure owner.
Selangor’s digital future
DxF comprises 30 initiatives across three implementation phases, but its immediate priority is fixing what Fahmi calls “legacy issues”, which is the inconsistent digital infrastructure that currently undermines the state’s tech ambitions. He offers a stark assessment of the major cloud service providers (CSPs) investing in Selangor which bypass the local fibre and internet infrastructure entirely.
“We are finding that foreign CSPs that are deploying data centres in Selangor sometimes have to create their own fibre network because what we have may not be reliable or up to their standards,” Fahmi says.
He provides details on the extent of redundancy these global players require. “There are not two, not three, but seven independent paths connecting the data centre, and every single path is built based on international standards,” he points out.
“What does that tell you? Do we have adequate standards for fiberisation in the state? If we do, then why do they have to build their own fibre networks?”
To close this gap, the state government is moving from connectivity standards in service level agreements (SLAs), which offer “best effort” connectivity, to service level guarantees (SLGs). Fahmi stresses that artificial intelligence-driven traffic systems and autonomous manufacturing cannot function on consumer-grade internet stability.
“You cannot run the state’s critical infrastructure and operations based on SLAs or best effort. If things don’t work, planes may fall from the sky or autonomous vehicles may malfunction,” he says.
The state is implementing strict enforcement. A subcommittee working with the Malaysian Communications and Multimedia Commission (MCMC) is monitoring tower operations for frequency and power consumption compliance.
“We have the right to pull the operating permit and cut off power,” says Fahmi, referring to operators who breach safety limits to cut costs. “If you’re operating illegally, we can shut it down.”
Under DxF, the state government is building its own dark fibre network and will appoint a state entity to deploy and manage fibre infrastructure specifically for public-sector connectivity.
When tabling the state’s Budget 2026, Selangor Menteri Besar Datuk Seri Amirudin Shari proposed the development of a dark fibre network master plan, with RM1 million allocated to the initial planning. The 889km network is aimed at providing the state with its own high-speed fibre-optic infrastructure.
Fahmi says the master plan does not mean the state government is trying to be a reseller of unused bandwidth but rather to create its own stable fibre network. “I think the way for us to become a subscriber of unused capacity is possible, but we shouldn’t be a reseller of the unused telco dark fibre facility because who will we resell it to?”
Selangor is positioning itself as the primary use case for the federal government’s upcoming sovereign private cloud rollout.
“The prime minister announced that for 2026, RM2 billion will be allocated to MCMC to build a sovereign private cloud infrastructure. We have been in discussions with MCMC, and we want to be the state that takes on that use case,” says Fahmi.
The state government is exploring graphics processing units-as-a-service (GPUaaS) and evaluating whether offering high-performance computing could potentially be a new revenue stream. It is also looking at implementing practical bureaucratic applications, such as deploying local large language (LLM) models to automate administration.
“The simplest example is to have an LLM that can be used for a speech-to-text agent so that your meeting minutes get recorded quickly. And it has to be the same model for local councils as well as state, federal and other agencies,” says Fahmi.
He says these models could be trained to include institutional knowledge, for example, a ChatGPT model that understands the nuances of state fatwas or religious rulings. The technology could also be expanded to include other public services that require specific local data.
Another application in the works is integrating topography mapping with real-time weather models to provide street-level flood predictions.
“You have your global macro weather model, but how do you bring it down to the micro level to know if a certain road in Petaling Jaya will flood? To achieve that flood-predictive model, we have to build the city’s digital twin,” he notes.
The most significant policy shift lies in the state’s philosophy on its planned super app. Fahmi explicitly rejects the commercial models of private e-wallets.
“People say that it must be like Touch ’n Go. I disagree. The whole country has thought of this in the wrong way,” he says.
He stresses that government apps should not be profit centres but cost centres designed for service delivery, funded by the state to reduce physical overheads.
“Even if it doesn’t make money, it has to be paid by the government. If you do the maths properly, I think it’s way cheaper to manage an app than to manage 12 citizen counters across the state,” he points out.
“We are a state government thinking about what we want to have in the next e-wallet. But the government app is not necessarily an e-wallet. What you need is a service delivery channel.”
To drive adoption without commercial marketing, the state government plans to leverage its regulatory power. Fahmi has proposed embedding the requirement for using the app directly into business licences.
“Imagine using all the powers we already have. We issue licences to SME businesses. Why can’t we make it a requirement that, as part of the licences, they must promote or list their businesses on the app?
“Doesn’t that help SMEs? It does. We can do all of this, but why aren’t we doing it yet? We keep saying SMEs are not big enough, when the real issue is that we have not required them to take part and use the right tools.”
The state’s interventionist approach stems from a fundamental rejection of what Fahmi calls the “Western model”, relying on private-sector innovation to solve public infrastructure problems. He points to the disparity between the wealth of Silicon Valley tech giants and the state of public infrastructure in the US.
“You think Apple is the biggest company, Nvidia is the biggest tech company. Go to California and see the condition of the roads there. Don’t tell me they don’t have potholes, they don’t have jobless people on the streets,” he says.
Instead, Selangor aims to emulate the top-down, government-led strategies of Singapore and China. “China took that model from Singapore five years ago. Where is China today? Sixty per cent of patents in the world on AI come from China,” he points out.
Fahmi argues that capitalist-driven innovation, while creating individual success stories, fails to deliver holistic benefits. “Forget the SMEs leading the way. We as the state government have to dictate. We have to write the storyline for how industrialisation should be done. Do not take the capitalistic model of the West. It doesn’t work anymore.”
Semiconductor investments
Complementing the digital infrastructure is the state’s aggressive entry into semiconductor capital. Ng says the state government aims to go beyond its initial RM100 million venture capital fund to support local IC design firms to not only build a base here but to successfully market their products.
Acknowledging that deep-tech firms often run pre-profit for years due to research and development (R&D) costs, the state is engaging with the Securities Commission Malaysia to draft new initial public offering (IPO) guidelines suitable for these companies.
“We are not expecting them to become unicorns overnight. That is not realistic,” says Ng, adding that the state is targeting sustainable “centaur” companies, or those with valuations of more than US$100 million, that can list within three to five years.
Centaur companies are start-ups valued at more than US$100 million but less than US$1 billion. These companies have proven business models and are growing steadily.
Following the launch of IC Design Park 2 in Cyberjaya on Nov 6 last year, instructions have been issued to project manager Selangor Information Technology and Digital Economy Corporation (Sidec) to explore establishing a third park in the next 12 months, says Ng.
While Park 1 focused on general IC design and Park 2 on testing and ecosystem support, Park 3 will target specific high-value niches. “Park 3 is probably going to be more focused on specific sectors, such as aerospace and automotive,” he says.
Addressing the talent crunch, specifically the immediate need for a total of 200 engineers at Park 1 and Park 2, Ng has proposed a radical expansion of Selangor’s role in recruitment. Moving beyond job matching, he envisions a concierge division that acts as a full-service headhunter and lifestyle consultant for expatriates and the returning Malaysian diaspora.
“I told Sidec that we may need to consider setting up another special division for this recruitment exercise. We are providing everything, including a recruitment exercise and real estate consultancy,” he says.
Selangor’s role is to remove friction, handling everything from visa processing to finding rental homes, says Ng. “We help them to handle all these hassles. You just come here and work.”
Addressing concerns that the influx of data centres contributes little to the local economy, he stresses that data centres built in the state must be part of the Local Economy Development (LED) policy. Formerly known as local content, the policy maintains a strict mandate that 30% of data centre capital expenditure must flow to local vendors.
“If we do not have any control or incentive, you are not bringing much benefit to our economy,” says Ng. Under LED, local SMEs must be integrated into the supply chain for cooling systems, server racks and maintenance, for example.
He stresses that the state is focused on realising its aspiration to be part of the higher-end value chain for semiconductors, adding that it “can’t afford to fail”.
Bets on autonomous technology and underserved global markets begin to pay off
When Selangor launched its first IC Design Park in August 2024, the move looked ambitious for a state that had been known more for services than complex chip designs.
With a second park now operational and 17 companies moving between the two facilities, the strategy has become clearer. IC Design Park 1 now houses 205 engineers, with plans to reach 400 across both parks by end-2025 as companies expand operations.
Rather than competing head on with global foundries wedged between the heightened US-China trade war, Selangor is carving out a niche in autonomous systems, robotics, drones and specialised data centres, where demand is rising but supply remains uneven.
Selangor Information Technology and Digital Economy Corporation (Sidec) chief executive Yong Kai Ping says this focus is deliberate by targeting inference applications for specific verticals compared to traditional semiconductor manufacturing.
“We should go into an industry that is less competitive but has a lot of usage, and aim for the automotive industry, humanoid robots and also drones and low-altitude aerospace engineering. These sectors already have a lot of proven use cases,” he says.
The strategy anticipates locally designed chips reaching production by 2027-2029.
While Western semiconductor firms focus on advanced markets, emerging economies across South America, South Africa, Gulf Cooperation Council countries and former Soviet states represent underserved opportunities, says Yong.
Citing Brazil as an example, he notes that the country requires about 20,000 drones for its agricultural sector, pointing to the information gathered during his visit to the country for the 2024 G20 Rio de Janeiro Summit.
These are markets that value affordability and reliability over cutting-edge performance, says Yong, adding that this approach leverages existing industrial strengths in the Klang Valley.
The automotive sector produced more than 700,000 vehicles annually from 2023 to 2025. Chinese electric vehicle (EV) manufacturers are also establishing local assembly operations in the country.
Launched on June 5 last year, the Selangor Aero Park, which brings together all aerospace activities in the state, targets 73% of national aerospace economic activity by the end of the decade.
Yong believes that by 2027 and 2028, more vehicles are expected to feature Level 3 or Level 4 autonomous capabilities, creating immediate domestic demand for advanced driver-assistance system (ADAS) chips and automotive-grade semiconductors.
The strategy includes pushing for a 30% local content requirement in data centres and automotive manufacturing, which aligns with the national industrial policy. This is aimed at compelling multinational corporations with operations in Malaysia to source from local parties, potentially relocating original design manufacturing (ODM) facilities to serve those customers.
“When we say there’s a local content policy in data centres in Selangor, then the Malaysia branch will be chosen first and they will then bring their ODM and more manufacturing facilities to Malaysia because their customers are already here,” says Yong.
A significant part of the strategy centres on talent. For decades, Malaysia’s engineering graduates have left for other talent-hungry markets, especially Singapore, enticed by higher salaries and more defined career pathways, but that exodus is slowing, he believes.
Through the Advanced Semiconductor Academy of Malaysia (Asem), Sidec now runs intensive pre-employment training and works directly with more than 20 Malaysian universities.
Sidec and Asem have become “the sole Malaysia-based group on the ground working with semiconductor talents”, and this presence has convinced many graduates to stay, says Yong.
The academy provides pre-job training, on-the-job training and job placement, creating what he describes as an “end-to-end service”. It had trained 300 students with full job placement as at 2025. It targets 600 annually as its expanded facilities come on stream.
With entry salaries of RM5,000 to RM6,000 for undergraduates, the ecosystem is producing “the highest starting salary bracket” in Malaysia’s semiconductor sector, says Yong. While it is below that offered in Singapore, the competitive local salary combined with the lower cost of living here make it appealing, he explains.
Yong disagrees that Selangor competes with Penang for talent. “That is wrong. The real contest has always been with Singapore.” By offering direct pathways to chip design roles, the Selangor government has “stopped the haemorrhaging of talent to Singapore”, he adds.
The geographical advantage is fundamental as 90% of engineering graduates come from universities in the Klang Valley, says Yong
Where Penang serves export-oriented multinational corporations, Selangor companies target local automotive, data centre and aerospace markets, which have different business models requiring different skill sets.
On capacity building, Yong is positive that senior Malaysian experts who left for China, Taiwan and Singapore are reconsidering their return as multinational corporations establish operations in Selangor, allowing them to lead those facilities. Some even return to set up their own companies here, he says.
“You have one senior from overseas leading 10 junior local engineers. After three to four years, all 10 junior engineers will become seniors as well,” he notes.
This mirrors how South Korea, China and India bootstrapped their semiconductor expertise by recruiting veterans from Taiwan Semiconductor Manufacturing Co Ltd (TSMC).
Selangor is expanding its collaboration with universities in India and China by offering an annual tuition of less than RM15,000, which is more accessible than that offered by many private colleges in Malaysia, says Yong. “The best teaching talent in the semiconductor industry are in China, India and Taiwan, not the West,” he adds.
The aim is to push Malaysian engineers towards master’s degrees, noting that nearly 90% of the workforce at top semiconductor exporters hold postgraduate degrees.
Selangor’s RM100 million venture capital fund is another anchor, with Sidec co-investing alongside three to four regional venture capitalists (VCs) per deal, lifting its total investment portfolio to about RM500 million. Its partners include regional funds as well as investors from Taiwan and Japan with an established track record in the semiconductor field.
The fund is split into two tranches. The first has already been deployed across seven to eight companies and the second is now being prepared. Going forward, Sidec’s investments will be closely aligned with the Arm initiative.
Last March, the federal government signed a US$250 million, 10-year partnership with UK-based Arm Holdings. The deal gives Malaysian companies access to seven Arm chip design blueprints and includes a commitment to train 10,000 local engineers.
The government expects the collaboration to accelerate domestic chip design capabilities, with the goal of producing Malaysian-made graphics processing units (GPUs) in five to 10 years and creating 10 home-grown chip companies, with each generating an annual revenue of US$1.5 billion to US$2 billion.
“The companies that are getting Arm chips, I think we will invest in a large number of them,” says Yong, adding that an IC design firm is preparing for an IPO, which could pave the way for Malaysia’s first semiconductor unicorn and validate the national target of having 10 semiconductor companies with a valuation of more than RM1 billion.
Save by subscribing to us for your print and/or digital copy.
P/S: The Edge is also available on Apple's App Store and Android's Google Play.
The content is a snapshot from Publisher. Refer to the original content for accurate info. Contact us for any changes.
Comments