Malaysia 5G and Edge
Network Technology Landscape

Malaysia's 5G networks are Non-Standalone (NSA). In NSA architecture, 5G radio antennas transmit data at 5G speeds, but all control signalling and session management routes through the existing 4G Evolved Packet Core. The practical result is faster downloads — but no network slicing, no sub-millisecond latency guarantees, and no ability to offer the dedicated network segments that enterprise customers in manufacturing, logistics, and healthcare would actually pay a premium for. NSA is a migration step, not a destination.

U Mobile is the one operator that has moved furthest. Its Ultra5G Advanced Network, which expanded outdoor coverage to Bangi, Putrajaya, and Petaling Jaya in late 2025, operates a hybrid configuration combining SA and NSA modes.[Telecompaper] This is not a full commercial SA launch — it is an architecture test at scale. No other Malaysian operator has confirmed equivalent SA trials in any public filing or announcement reviewed for this report. Maxis, CelcomDigi, YES, and TM all remain on NSA with no publicly committed SA migration timelines.

The ASEAN regional pattern confirms this is not a Malaysia-specific lag. Across Asia Pacific, NSA dominates public and private 5G deployments because device ecosystem maturity for SA remains uneven and spectrum readiness varies significantly by market.[MarketsandMarkets] China is the only APAC market where SA deployments are at scale. The implication for Malaysia: SA migration is a 2027–2028 story at the earliest, and any enterprise use case that depends on SA capabilities — private network slicing, ultra-reliable low latency — cannot be sold commercially today.

No Malaysian telco — Maxis, CelcomDigi, U Mobile, YES, or TM — has publicly named its 5G radio access network vendor in any regulatory filing, investor presentation, or press release that appears in Tier 1 or Tier 2 sources reviewed for this report. This is unusual. In comparable markets — Indonesia, Thailand, the Philippines — at least partial vendor relationships are disclosed through procurement notices, partnership announcements, or infrastructure-sharing agreements. In Malaysia, the opacity is near-total.

The opacity has a structural explanation. Malaysia's original DNB (Digital Nasional Berhad) single-wholesale-network model, which contracted Ericsson as its primary RAN vendor in 2021, was substantially restructured from 2023 onward as operators were permitted to build their own 5G infrastructure. The transition from a single shared network to a competitive multi-operator buildout created a period of contract renegotiation and new procurement that is ongoing — and operators have chosen not to disclose which vendors they are working with in this new structure.

The global vendor context is relevant but cannot be applied directly to Malaysia without named evidence. Ericsson, Huawei, Nokia, ZTE, and Samsung are the five credible 5G RAN suppliers globally. Huawei faces procurement restrictions in some markets due to security concerns, which has influenced vendor decisions in Singapore, Australia, and parts of Europe — but Malaysia has not issued equivalent guidance, and Huawei remains an active commercial presence in Southeast Asia. Without operator disclosure, no vendor share can be stated with any confidence. Any figure cited in this space without a named Malaysian source is an invention.

MCMC has allocated 700 MHz and 3.5 GHz spectrum bands for 5G deployment in Malaysia — the 700 MHz band providing wide-area coverage and indoor penetration, and 3.5 GHz the high-capacity mid-band used for urban 5G throughput. Both assignments happened in 2022–2023. What has not happened is any official published timeline for 2G or 3G network decommissioning by any operator. No MCMC directive requiring shutdown by a specific date has been identified in any source reviewed for this report.[MCMC]

The absence of a shutdown mandate creates a strategic ambiguity. Operators maintaining 2G and 3G infrastructure are paying ongoing network operating costs for a diminishing subscriber base — but they are also preserving coverage in rural areas where 4G and 5G signals are not yet reliable. The spectrum freed by 2G shutdown (primarily 900 MHz and 1800 MHz bands) would meaningfully improve LTE and 5G indoor coverage if refarmed, but the timing of that refarming is operator-discretionary until MCMC sets a regulatory endpoint. In comparable regional markets, Thailand completed its 2G shutdown in 2022 and the Philippines in 2023 — Malaysia has not confirmed an equivalent date.

The MyTMAP2030 roadmap references digital infrastructure ambitions through 2030 but sets no binding 2G/3G exit dates in the publicly available version of the document.[MCMC] This is a gap in regulatory clarity that affects investment planning: vendors bidding for LTE refarming contracts and enterprise customers planning private 5G in rural areas both need to know when legacy spectrum becomes available.

The Asia Pacific private 5G market — dedicated 5G networks for enterprise campuses, factories, ports, and campuses — is growing at 39% annually and will reach USD 8.0B by 2030 from USD 1.5B in 2025.[MarketsandMarkets] Manufacturing and logistics are the primary drivers, because private 5G enables wireless machine control, automated guided vehicles, and real-time quality inspection at latencies that Wi-Fi cannot reliably guarantee. Malaysia's manufacturing base — electronics, semiconductors, automotive — is structurally well-positioned to absorb this technology.

What is actually deployed in Malaysia sits at a much earlier stage. U Mobile's partnership with the Malaysia Retail Chain Association (MRCA) in 2025 covers 5G-enabled smart retail applications: cashless payments, inventory management, and customer analytics at sites including Berjaya Times Square and the Penang Bridge 1 corridor.[Developing Telecoms] Maxis has a separate MRCA partnership targeting AI computer vision, extended reality, and supply chain analytics in retail environments.[Developing Telecoms] These are real commercial engagements — but they are retail pilots, not industrial private networks. No Malaysian manufacturer, port, or airport has announced a private 5G deployment with a named operator and a named technology stack.

The gap between the market growth rate and Malaysia's current deployment stage is not a failure — it is a timing indicator. The technology and the demand are converging. The missing ingredient is 5G Standalone architecture, which is required for the network slicing that makes private 5G financially distinct from a dedicated Wi-Fi 6 deployment. Until operators launch SA cores, private 5G in Malaysia is commercially positioned but technically constrained.

Multi-access Edge Compute (MEC) places processing power physically close to the radio access network — typically inside or adjacent to a base station or a telco data centre — so that applications get single-digit millisecond response times instead of the 30–80ms round trip to a central cloud region. This matters for autonomous systems, real-time video analytics, and industrial control applications. Without MEC, 5G's latency advantage is theoretical rather than delivered.

No Malaysian telco has announced a named MEC deployment with AWS, Microsoft Azure, or Google Cloud as of Q1 2026. AWS Wavelength — Amazon's operator-embedded edge product — is live in Japan, South Korea, and the United States. Azure Private MEC has named deployments in India and Japan. Google Distributed Cloud Edge has pilots in Singapore. None of these programmes have a confirmed Malaysian telco partner in any source reviewed. Malaysian hyperscaler presence is primarily expressed through cloud region infrastructure: AWS (Kuala Lumpur region live since 2024), Azure (Malaysia region live since 2023), and Google Cloud (Malaysia region announced 2024) — but cloud regions and edge compute are different products serving different latency requirements.

The implication is that Malaysia's 5G edge story is at minimum 18–24 months behind comparable deployments in Japan and South Korea, and roughly 12 months behind Singapore. The demand signal from Malaysia's manufacturing and logistics sector is strong enough to attract this investment — but it will require SA core maturity at the operator level before hyperscalers can build commercially viable edge products on top of Malaysian 5G.

The 2025 Malaysia Artificial Intelligence Research and Report found that 84% of Malaysian enterprises — across all sectors — remain in the exploration stage for AI.[Malaysia AI Report] This is the baseline against which telco AI deployment must be read. Network operations AI, predictive maintenance systems, and autonomous network management platforms are all commercially available from vendors including Ericsson (AI-native RAN), Nokia (AVA platform), and Huawei (iMaster NCE) — but no Malaysian operator has disclosed a named production deployment of any of these systems in any source reviewed for this report.

The distinction between exploration and production matters for network operations specifically. AI-assisted fault detection and predictive maintenance in telco networks can reduce mean-time-to-repair by 30–50% according to vendor case studies from deployed networks in Europe and South Korea — but those gains are only available once the AI system is trained on real network data at production scale. Pilots running on synthetic or limited datasets produce vendor demonstrations, not operational improvements. The absence of named production deployments in Malaysia means these gains are not yet flowing to any Malaysian operator's network quality or cost base.

The one named deployment adjacent to AI in Malaysian 5G is U Mobile and Maxis's retail partnerships with MRCA, which include AI computer vision for store analytics and customer behaviour tracking.[Developing Telecoms] These are customer-facing AI applications sitting on top of 5G connectivity — not network operations AI. The distinction matters: network AI improves the infrastructure; retail AI monetises it. Malaysia has early movement on the monetisation side and near-zero confirmed production on the infrastructure side.

The next 24–36 months in Malaysia's 5G market will be determined by three compounding decisions: when operators migrate to Standalone cores, which operator secures the first credible hyperscaler MEC partnership, and whether MCMC sets a binding 2G/3G shutdown date that frees mid-band spectrum for refarming. Each of these is independently valuable — together they define whether Malaysia becomes a regional reference market for enterprise 5G or watches Singapore extend its lead.

U Mobile's hybrid SA/NSA trial positions it as the most likely first mover on SA launch, which would give it a genuine technology differentiation that NSA peers cannot immediately replicate. But U Mobile is the smallest of the five operators by subscriber base — SA leadership does not automatically translate to enterprise market share without the sales infrastructure and partner ecosystem to commercialise it. The operator that pairs SA architecture with a named hyperscaler edge partnership first will have the most complete enterprise proposition in the market.

The base case — the most probable outcome given current trajectory — is that one operator (most likely U Mobile or Maxis) achieves a commercial SA launch by end-2027, hyperscaler MEC partnerships are announced but not at production scale until 2028, and Malaysian enterprise private 5G grows materially but remains concentrated in retail and light industrial use cases rather than the heavy manufacturing applications that drive the highest revenue per site. The structural driver of this timeline is not technology readiness — the technology exists — but the commercial decision-making cycle inside operators that have competing capital priorities and no regulatory deadline forcing SA migration.