AI is reshaping the tech infrastructure landscape, demanding a fundamental rethinking of organisational infrastructure strategies. Traditional infrastructure, once sufficient, now struggles to keep pace with the immense scale and complexity of AI workloads. To meet these demands, organisations are turning to high-performance computing (HPC) solutions, leveraging powerful GPUs and specialised accelerators to handle the computationally intensive nature of AI algorithms.
Real-time AI applications, from fraud detection to autonomous vehicles, require lightning-fast processing speeds and low latency. This is driving the adoption of high-speed networks and edge computing, enabling data processing closer to the source and reducing response times. AI-driven automation is also streamlining infrastructure management, automating tasks like network provisioning, security monitoring, and capacity planning. This not only reduces operational overhead but also improves efficiency and frees up valuable resources.
Ecosystm analysts Darian Bird, Peter Carr, Simona Dimovski, and Tim Sheedy present the key trends shaping the tech infrastructure market in 2025.
Click here to download ‘Building the AI Future: Top 5 Infra Trends for 2025’ as a PDF
1. The AI Buildout Will Accelerate; China Will Emerge as a Winner
In 2025, the race for AI dominance will intensify, with Nvidia emerging as the big winner despite an impending AI crash. Many over-invested companies will fold, flooding the market with high-quality gear at bargain prices. Meanwhile, surging demand for AI infrastructure – spanning storage, servers, GPUs, networking, and software like observability, hybrid cloud tools, and cybersecurity – will make it a strong year for the tech infrastructure sector.
Ironically, China’s exclusion from US tech deals has spurred its rise as a global tech giant. Forced to develop its own solutions, China is now exporting its technologies to friendly nations worldwide.
By 2025, Chinese chipmakers are expected to rival international peers, with some reaching parity.
2. AI-Optimised Cloud Platforms Will Dominate Infrastructure Investments
AI-optimised cloud platforms will become the go-to infrastructure for organisations, enabling seamless integration of machine learning capabilities, scalable compute power, and efficient deployment tools.
As regulatory demands grow and AI workloads become more complex, these platforms will provide localised, compliant solutions that meet data privacy laws while delivering superior performance.
This shift will allow businesses to overcome the limitations of traditional infrastructure, democratising access to high-performance AI resources and lowering entry barriers for smaller organisations. AI-optimised cloud platforms will drive operational efficiencies, foster innovation, and help businesses maintain compliance, particularly in highly regulated industries.
3. PaaS Architecture, Not Data Cleanup, Will Define AI Success
By 2025, as AI adoption reaches new heights, organisations will face an urgent need for AI-ready data, spurring significant investments in data infrastructure. However, the approach taken will be pivotal.
A stark divide will arise between businesses fixated on isolated data-cleaning initiatives and those embracing a Platform-as-a-Service (PaaS) architecture.
The former will struggle, often unintentionally creating more fragmented systems that increase complexity and cybersecurity risks. While data cleansing is important, focusing exclusively on it without a broader architectural vision leads to diminishing returns. On the other hand, organisations adopting PaaS architectures from the start will gain a distinct advantage through seamless integration, centralised data management, and large-scale automation, all critical for AI.
4. Small Language Models Will Push AI to the Edge
While LLMs have captured most of the headlines, small language models (SLMs) will soon help to drive AI use at the edge. These compact but powerful models are designed to operate efficiently on limited hardware, like AI PCs, wearables, vehicles, and robots. Their small size translates into energy efficiency, making them particularly useful in mobile applications. They also help to mitigate the alarming electricity consumption forecasts that could make widespread AI adoption unsustainable.
Self-contained SMLs can function independently of the cloud, allowing them to perform tasks that require low latency or without Internet access.
Connected machines in factories, warehouses, and other industrial environments will have the benefit of AI without the burden of a continuous link to the cloud.
5. The Impact of AI PCs Will Remain Limited
AI PCs have been a key trend in 2024, with most brands launching AI-enabled laptops. However, enterprise feedback has been tepid as user experiences remain unchanged. Most AI use cases still rely on the public cloud, and applications have yet to be re-architected to fully leverage NPUs. Where optimisation exists, it mainly improves graphics efficiency, not smarter capabilities. Currently, the main benefit is extended battery life, explaining the absence of AI in desktop PCs, which don’t rely on batteries.
The market for AI PCs will grow as organisations and consumers adopt them, creating incentives for developers to re-architect software to leverage NPUs.
This evolution will enable better data access, storage, security, and new user-centric capabilities. However, meaningful AI benefits from these devices are still several years away.
2024 was a year marked by intense AI-driven innovation. While the hype surrounding AI may have reached a fever pitch, the technology’s transformative potential is undeniable.
The growing interest in AI can be attributed to several factors: the democratisation of AI, with tools and platforms now accessible to businesses of all sizes; AI’s appeal to business leaders, offering actionable insights and process automation; and aggressive marketing by major tech companies, which has amplified the excitement and hype surrounding AI.
2025 will be a year defined by AI, with its transformative impact rippling across industries. However, other geopolitical and social factors will also significantly shape the tech landscape.
Ecosystm analysts Achim Granzen, Alan Hesketh, Audrey William, Clay Miller, Darian Bird, Manish Goenka, Richard Wilkins, Sash Mukherjee, Simona Dimovski, and Tim Sheedy present the key trends and disruptors shaping the tech market in 2025.
Click here to download ‘Key Tech Trends & Disruptors in 2025’ as a PDF
1. Quantum Computing Will Drive Major Transformation in the Tech Industry
Advancements in qubit technology, quantum error correction, and hybrid quantum-classical systems will accelerate breakthroughs in complex problem-solving and machine learning. Quantum communications will revolutionise data security with quantum key distribution, providing nearly unbreakable communication channels. As quantum encryption becomes more widespread, it will replace current cryptographic methods, protecting sensitive data from future quantum-enabled attacks.
With quantum computing threatening encryption standards like RSA and ECC, post-quantum encryption will be critical for data security.
While the full impact of quantum computers is expected within the next few years, 2025 will be pivotal in the transition toward quantum-resistant security measures and infrastructure.
2. Many Will Try, But Few Will Succeed as Platform Companies
Hypergrowth occurs when companies shift from selling products to becoming platform providers. Unlike traditional businesses, platforms don’t own inventory; their value lies in proprietary data and software that connect buyers, sellers, and consumers. Platforms disrupt industries and often outperform legacy businesses, with examples like Uber, Amazon, and Meta, and disruptors like Lemonade in insurance and Wise in international funds transfer.
In 2025, many companies will aim to become platform businesses, with AI seen as a key driver.
They will begin creating platforms and building ecosystems around them – some within existing brands, others launching new ones or even new subsidiaries to seize this opportunity.
3. A Trans-Atlantic Divide Will Emerge in AI Regulation
The EU is poised to continue its rigorous approach to AI regulation, emphasisng ethical considerations and robust governance. This is evident in the recent AI Act, which imposes stringent guidelines and penalties for violations. The EU’s commitment to responsible AI development is likely to lead to a more cautious and controlled innovation landscape.
In contrast, the US, under a new administration, may adopt a more lenient regulatory stance towards AI. This shift could accelerate innovation and foster a more permissive environment for AI development. However, it may also raise concerns about potential risks and unintended consequences.
This divergence in regulatory frameworks could create significant challenges for multinational companies operating in both regions.
4. The Rise of AI-Driven Ecosystem Platforms Will Shape Tech Investments
By 2025, AI-driven ecosystem platforms will dominate tech investments, fueled by technological convergence, market efficiency demands, and evolving regulations. These platforms will integrate AI, IoT, cloud, and data analytics to create seamless, predictive ecosystems that transcend traditional industry boundaries.
Key drivers include advancements in AI, global supply chain disruptions, and rising ESG expectations. Regulatory shifts, such as the EU’s AI Act, will further push for compliant, ethical platforms emphasising transparency and accountability.
For businesses, this shift redefines technology as interconnected ecosystems driving efficiency, innovation, and customer value.
5. AI-Powered Data Fabrics Will be the Foundation for Data-Driven Success
In 2025, AI-powered data fabrics will become a core technology for large organisations.
They will transition from basic data management tools to intelligent systems that deliver value across the entire data lifecycle. Organisations will finally be able to get control of their data governance.
AI’s enhanced role will automate essential data functions, including intelligent data integration and autonomous connection to diverse data sources. AI will also enable proactive data quality management, predicting and preventing errors for improved reliability. AI-driven data fabrics will also offer automated data discovery and mapping, dynamic data quality and governance, intelligent data integration, and enhanced data access and delivery.
6. Focus Will Shift From AI Models to Intelligence Gaps & Performance
While many organisations are investing in AI, only those that started their transformation in 2024 are truly AI-led. Most have become AI-driven through embedded AI in enterprise systems as tech providers continue to evolve their offerings. However, these multi-vendor environments often lack synergy, creating gaps and blind spots.
In 2025, organisations will pause their investments to assess AI capabilities and identify these gaps.
Once they pinpoint the blind spots, investments will refocus not on new AI models, but on areas like model orchestration to manage workflows and ensure peak performance; vendor management to establish unified governance frameworks for flexibility and compliance; and eventually automated AI lifecycle management, with centralised inventories and monitoring to track performance and detect issues like model drift.
7. Specialised Small Language Models Will Gain Traction
GenAI, driven by LLMs, has dominated the spotlight, fueling both excitement and concerns about AI. However, LLM-based GenAI is entering a phase of diminishing returns, both in terms of individual model capabilities and the number of available models. Only a few providers will have the resources to develop LLMs, focusing on a limited number of models.
This will see the increased popularity of small language models (SLMs), that are tailored for a specific purpose, use case, or environment. These models will be developed by startups, organisations, and enterprises with deep domain knowledge and data. They will be fully commercialised driving narrow but distinct ROI.
There will be an increased demand for GPU-as-a-service and SLM-as-a-service, and the platforms which can support these.
8. Multi-agent AI Systems Will Help Manage Complexity and Collaboration
Isolated AI tools that can perform narrow tasks lack the adaptability and coordination required for real-time decision-making. Multi-agent systems, in contrast, consist of decentralised agents that collaborate, share information, and make independent decisions while working toward a common goal. This approach not only improves efficiency but also enhances resilience in rapidly changing conditions.
Early use cases will be in complex environments that require cooperation between multiple stakeholders.
Multi-agent systems will optimise logistics by continuously analysing disruptions and dynamically balancing supply and demand in energy grids. These multi-agent systems will also operate in competitive modes, such as algorithmic trading, ad auctions, and ecommerce recommender systems.
9. Super Apps Will Expand into Rural & Underserved Markets in Asia Pacific
Super apps are set to reshape rural economies, fueled by increased internet access, affordable tech, and heavy government investment in digital infrastructure. Their localised, all-in-one services unlock untapped potential in underserved regions, fostering inclusivity and innovation.
By 2025, super apps will deepen their reach across Asia, integrating communication, payments, and logistics into seamless platforms.
Leveraging affordable mobile devices, cloud-native technologies, and localised services, they will penetrate rural and underserved areas with tailored solutions like agricultural marketplaces, local logistics, and expanded government services. Enterprises investing in agile cloud infrastructure will drive this evolution, bridging the digital divide, boosting economic growth, and enhancing user experiences for millions.
10. Intense Debates Over Remote vs. In-Office Work Will Persist in Asia Pacific
Employers in Asia Pacific will enforce stricter return-to-office policies, linking them to performance metrics and benefits to justify investments in physical spaces and enhance workforce productivity.
However, remote collaboration will remain integral, even for in-office teams.
The push for human-centred tech will grow, focusing on employee well-being and flexibility through AI-powered tools and hybrid platforms. Companies will prioritise enhancing employee experiences with personalised, adaptable workspaces, while office designs will increasingly incorporate biophilic elements, blending nature and technology to support seamless collaboration and remote integration.
GenAI has taken the world by storm, with organisations big and small eager to pilot use cases for automation and productivity boosts. Tech giants like Google, AWS, and Microsoft are offering cloud-based GenAI tools, but the demand is straining current infrastructure capabilities needed for training and deploying large language models (LLMs) like ChatGPT and Bard.
Understanding the Demand for Chips
The microchip manufacturing process is intricate, involving hundreds of steps and spanning up to four months from design to mass production. The significant expense and lengthy manufacturing process for semiconductor plants have led to global demand surpassing supply. This imbalance affects technology companies, automakers, and other chip users, causing production slowdowns.
Supply chain disruptions, raw material shortages (such as rare earth metals), and geopolitical situations have also had a fair role to play in chip shortages. For example, restrictions by the US on China’s largest chip manufacturer, SMIC, made it harder for them to sell to several organisations with American ties. This triggered a ripple effect, prompting tech vendors to start hoarding hardware, and worsening supply challenges.
As AI advances and organisations start exploring GenAI, specialised AI chips are becoming the need of the hour to meet their immense computing demands. AI chips can include graphics processing units (GPUs), application-specific integrated circuits (ASICs), and field-programmable gate arrays (FPGAs). These specialised AI accelerators can be tens or even thousands of times faster and more efficient than CPUs when it comes to AI workloads.
The surge in GenAI adoption across industries has heightened the demand for improved chip packaging, as advanced AI algorithms require more powerful and specialised hardware. Effective packaging solutions must manage heat and power consumption for optimal performance. TSMC, one of the world’s largest chipmakers, announced a shortage in advanced chip packaging capacity at the end of 2023, that is expected to persist through 2024.
The scarcity of essential hardware, limited manufacturing capacity, and AI packaging shortages have impacted tech providers. Microsoft acknowledged the AI chip crunch as a potential risk factor in their 2023 annual report, emphasising the need to expand data centre locations and server capacity to meet customer demands, particularly for AI services. The chip squeeze has highlighted the dependency of tech giants on semiconductor suppliers. To address this, companies like Amazon and Apple are investing heavily in internal chip design and production, to reduce dependence on large players such as Nvidia – the current leader in AI chip sales.
How are Chipmakers Responding?
NVIDIA, one of the largest manufacturers of GPUs, has been forced to pivot its strategy in response to this shortage. The company has shifted focus towards developing chips specifically designed to handle complex AI workloads, such as the A100 and V100 GPUs. These AI accelerators feature specialised hardware like tensor cores optimised for AI computations, high memory bandwidth, and native support for AI software frameworks.
While this move positions NVIDIA at the forefront of the AI hardware race, experts say that it comes at a significant cost. By reallocating resources towards AI-specific GPUs, the company’s ability to meet the demand for consumer-grade GPUs has been severely impacted. This strategic shift has worsened the ongoing GPU shortage, further straining the market dynamics surrounding GPU availability and demand.
Others like Intel, a stalwart in traditional CPUs, are expanding into AI, edge computing, and autonomous systems. A significant competitor to Intel in high-performance computing, AMD acquired Xilinx to offer integrated solutions combining high-performance central processing units (CPUs) and programmable logic devices.
Global Resolve Key to Address Shortages
Governments worldwide are boosting chip capacity to tackle the semiconductor crisis and fortify supply chains. Initiatives like the CHIPS for America Act and the European Chips Act aim to bolster domestic semiconductor production through investments and incentives. Leading manufacturers like TSMC and Samsung are also expanding production capacities, reflecting a global consensus on self-reliance and supply chain diversification. Asian governments are similarly investing in semiconductor manufacturing to address shortages and enhance their global market presence.
Japan is providing generous government subsidies and incentives to attract major foreign chipmakers such as TSMC, Samsung, and Micron to invest and build advanced semiconductor plants in the country. Subsidies have helped to bring greenfield investments in Japan’s chip sector in recent years. TSMC alone is investing over USD 20 billion to build two cutting-edge plants in Kumamoto by 2027. The government has earmarked around USD 13 billion just in this fiscal year to support the semiconductor industry.
Moreover, Japan’s collaboration with the US and the establishment of Rapidus, a memory chip firm, backed by major corporations, further show its ambitions to revitalise its semiconductor industry. Japan is also looking into advancements in semiconductor materials like silicon carbide (SiC) and gallium nitride (GaN) – crucial for powering electric vehicles, renewable energy systems, and 5G technology.
South Korea. While Taiwan holds the lead in semiconductor manufacturing volume, South Korea dominates the memory chip sector, largely due to Samsung. The country is also spending USD 470 billion over the next 23 years to build the world’s largest semiconductor “mega cluster” covering 21,000 hectares in Gyeonggi Province near Seoul. The ambitious project, a partnership with Samsung and SK Hynix, will centralise and boost self-sufficiency in chip materials and components to 50% by 2030. The mega cluster is South Korea’s bold plan to cement its position as a global semiconductor leader and reduce dependence on the US amidst growing geopolitical tensions.
Vietnam. Vietnam is actively positioning itself to become a major player in the global semiconductor supply chain amid the push to diversify away from China. The Southeast Asian nation is offering tax incentives, investing in training tens of thousands of semiconductor engineers, and encouraging major chip firms like Samsung, Nvidia, and Amkor to set up production facilities and design centres. However, Vietnam faces challenges such as a limited pool of skilled labour, outdated energy infrastructure leading to power shortages in key manufacturing hubs, and competition from other regional players like Taiwan and Singapore that are also vying for semiconductor investments.
The Potential of SLMs in Addressing Infrastructure Challenges
Small language models (SLMs) offer reduced computational requirements compared to larger models, potentially alleviating strain on semiconductor supply chains by deploying on smaller, specialised hardware.
Innovative SLMs like Google’s Gemini Nano and Mistral AI’s Mixtral 8x7B enhance efficiency, running on modest hardware, unlike their larger counterparts. Gemini Nano is integrated into Bard and available on Pixel 8 smartphones, while Mixtral 8x7B supports multiple languages and suits tasks like classification and customer support.
The shift towards smaller AI models can be pivotal to the AI landscape, democratising AI and ensuring accessibility and sustainability. While they may not be able to handle complex tasks as well as LLMs yet, the ability of SLMs to balance model size, compute power, and ethical considerations will shape the future of AI development.
