Digital infrastructure and AI are reshaping the global economy.At the centre of this shift lies a complex stack of systems and dependencies. AI runs on data centres, connects with enterprise platforms, and depends on energy infrastructure - all under tightening performance and sustainability demands.For nearly two decades, I've written about enterprise IT, data centres, AI, and cybersecurity, along with the operational realities behind them. Having worked across editorial, agency, and brand roles, I understand not just how technology works, but how narratives shape markets.In an era where generative AI can produce endless content, the advantage no longer lies in volume. It lies in perspective and authority.✨ How do you stand out when everyone is publishing?✨ How do you attract the right audience?✨ How do you build trust in a noisy market?That's where I come in.---What I bring:✅ Technical fluency from a prior career in IT.✅ 18 years of editorial judgement.✅ Deep familiarity with APAC data centre and digital infrastructure ecosystem.✅ Experience shaping narratives that resonate with technically literate audiences.---How we can work together: Work with me directly👉 Consulting on AI, data centre and technology trends.👉 Strategic content advisory and executive positioning.👉 Speaking, moderation and industry panels.Reach my audience👉 Strategic editorial collaborations.👉 Authority-building brand partnerships.---If you would like to read more of my content:🔸Follow me on LinkedIn.🔸Tech Stories Blog: https://www.techstoryteller.co🔸Weekly newsletter: https://www.techstories.co/updates🔸Clearly Tech Substack: https://www.clearlytech.co🔸WhatsApp Channel: https://whatsapp.com/channel/0029VbB6pYr9hXF32Ycd4t2f---You can contact me via LinkedIn message (no need to connect first) or email at: paul@techstoryteller.co

Paul Mah

STT GDC India last week announced the launch of its fourth data centre in Chennai, India, with phase one of the 45MW facility operational.

✅ Join my WhatsApp Channel (1-click): https://lnkd.in/gfZUWPSe

𝟭/ 𝗦𝗧𝗧 𝗚𝗗𝗖 𝗶𝗻 𝗜𝗻𝗱𝗶𝗮

Located in the SIPCOT IT Park, Siruseri, the STT Chennai 7 data centre is designed as a 45MW "AI-ready" facility. The first phase that is currently operational is 7.2MW.

STT GDC India also has 40MW already operational at its Chennai Ambattur campus, split between the STT Chennai 2 and STT Chennai 3 data centres.

Earlier this month, STT GDC India held the groundbreaking of STT Chennai 4, also in the Ambattur campus. According to the company website, it will offer 18MW of capacity.

𝟮/ 𝗣𝗼𝘀𝘀𝗶𝗯𝗹𝗲 𝗜𝗣𝗢 𝗺𝗼𝗼𝘁𝗲𝗱

In its press release, STT GDC India says it is evaluating options for future expansion in India, including a potential IPO in 1H 2027 to accelerate growth.

STT GDC holds a 74% stake in STT GDC India with Tata Communications holding the remainder - STT GDC in 2016 acquired a controlling interest in Tata Communications' data centre business in India.

A KKR-led consortium in February announced the full acquisition of STT GDC at an enterprise value of S$13.8 billion. (Read: https://lnkd.in/ggFvPASA)

𝟯/ 𝗜𝗻𝗱𝗶𝗮 𝗼𝗻 𝘁𝗵𝗲 𝗳𝗮𝘀𝘁 𝘁𝗿𝗮𝗰𝗸

As I noted in August last year, there's just 1.4 GW of data centres in operation across India today. That's a dismal figure given the fact that it is the world's most-populous nation with over 1.4 billion people.

The Indian government is making a concerted push to woo investors to get on the AI bandwagon and leapfrog ahead in the digital infrastructure race.

On 1 Feb, Finance Minister Nirmala Sitharaman announced that India will offer a 20-year tax holiday to overseas firms that offer global data centre services from the South Asian country.

Finally, India has other things going for it, including some 11.18GW of renewables at the moment and a push to develop more.

📆 𝗚𝗲𝘁 𝗺𝘆 𝗰𝗼𝗻𝘁𝗲𝗻𝘁

✅ Weekly newsletter: techstories.co/updates
✅ WhatsApp Channel (If 1-click doesn't work): techstories.co/wa

𝗣𝗵𝗼𝘁𝗼 𝗖𝗿𝗲𝗱𝗶𝘁: STT GDC. STT Chennai 7.

Helping growing tech companies with their content and strategy

Adani Group this week announced plans to invest as much as US$100 billion by 2035 to develop AI data centres in India.

Specifics are not provided, though the press release says it builds on AdaniConneX's existing 2GW national data centre roadmap and will take total planned target to 5GW.

AdaniConneX is a 50:50 joint venture by Adani Enterprises and US-based EdgeConneX, and is touted as India's largest hyperscale data centre operator.

✅ Join my WhatsApp Channel (one-click): https://lnkd.in/gfZUWPSe

𝟭/ 𝗔𝗜 𝗱𝗮𝘁𝗮 𝗰𝗲𝗻𝘁𝗿𝗲𝘀

The data centre expansion is anchored by partnerships with Google, Microsoft and Flipkart. According to Adani, this includes:

🔸Google for a 1GW AI data centre campus in Visakhapatnam, as well as additional campuses in Noida.

🔸The building of data centres spanning Hyderabad and Pune for Microsoft.

🔸Plans to develop a second AI data centre with India e-commerce firm Flipkart.

The Adani Group is also in discussion with other major players to build large scale data centre campuses across India.

𝟮/ 𝗘𝗻𝗱-𝘁𝗼-𝗲𝗻𝗱 𝘀𝘁𝗮𝗰𝗸

Adani Group chairman, Gautam Adani spoke of a five-layer AI stack, a framework that essentially covers apps, AI models, semiconductor chips, energy, and data centres.

The idea is to catalyse and leverage additional investments across server manufacturing and supply chains to make India an "AI-first superpower" as described by Adani's CFO.

Indeed, Adani is investing heavily to expand its renewable energy portfolio, which will include the world's largest battery energy storage systems (BESS).

𝟯/ 𝗗𝗲𝗺𝗮𝗻𝗱 𝗳𝗼𝗿 𝗱𝗮𝘁𝗮 𝗰𝗲𝗻𝘁𝗿𝗲𝘀

From a population and digitalisation standpoint, India currently does not have enough data centres to serve its own needs.

But Adani has clearly set its sights beyond traditional data centres and is looking to leapfrog ahead through multiple AI data centre deployments.

The India AI Impact Summit held in New Delhi this week has drawn attendees with around 10 heads of state and industry luminaries such as Alphabet's Sundar Pichai, OpenAI's Sam Altman, and Anthropic's Dario Amodei.

Where do you see India's data centre market heading?

📆 𝗚𝗲𝘁 𝗺𝘆 𝗰𝗼𝗻𝘁𝗲𝗻𝘁

✅ Weekly newsletter: techstories.co/updates
✅ WhatsApp Channel (If 1-click doesn't work): techstories.co/wa

𝗣𝗵𝗼𝘁𝗼 𝗖𝗿𝗲𝗱𝗶𝘁: AdaniConneX. AdaniConneX's Chennai 1 data centre in Chennai, India.

Singapore says it will impose PUE requirements on both old and new data centres with the upcoming Digital Infrastructure Act (DIA).

Speaking at the launch of a Centre of Excellence for Applied AI by Singtel's Digital InfraCo and Nvidia this morning, Senior Minister of State Tan Kiat How made two key announcements.

✅ Join my WhatsApp Channel (one-click): https://lnkd.in/gfZUWPSe

𝟭/ 𝗗𝗶𝗴𝗶𝘁𝗮𝗹 𝗜𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 𝗔𝗰𝘁 𝗰𝗼𝗺𝗶𝗻𝗴

The Digital Infrastructure Act (DIA), which is designed to enhance the security and resilience of critical digital infrastructure (CII) such as data centres, will be tabled in parliament later this year. 

First mentioned in January 2024, the DIA was originally scheduled to be introduced last year. It is not known why this was pushed back, though my personal speculation is that the DIA was meant to come after the DC-CFA2, which only kicked off in December 2025.

Another possibility could be the complexity of the legislation, or that CIIs in Singapore coming under state-sponsored attacks last year could have resulted in changes or new provisions being added. Regardless, it sounds like the DIA will be passed into law soon.

𝟮/ 𝗕𝗮𝘀𝗲𝗹𝗶𝗻𝗲 𝗲𝗻𝗲𝗿𝗴𝘆 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆

One new revelation that I wasn't aware before, is how DIA will also seek to establish "baseline energy efficiency requirements" for data centres in Singapore.

This includes both existing and new ones that will be built, which will see the imposition of PUE requirements, though no PUE figures were mentioned this morning.

In his speech, the minister also alluded to the systematic raising of the "energy efficiency and sustainability" of all data centres in Singapore to create "more headroom" to support growing digital demands here.

𝟯/ 𝗦𝘂𝘀𝘁𝗮𝗶𝗻𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗽𝗹𝗮𝘆

A regional data centre hub, Singapore has a plan to build a green energy hub and unlock sustainable data centre growth.

It is worth noting that the IMDA-BCA Green Mark for Data Centres 2024 standard imposes tighter requirements on how PUE is measured. This means operators should probably not count on being able to evade this requirement.

You can read more about this in "Singapore's audacious bet on sustainable data centres" here: go.clearlytech.co/k4Hj

I'll share more about the DIA when I learn more about it.

📆 𝗚𝗲𝘁 𝗺𝘆 𝗰𝗼𝗻𝘁𝗲𝗻𝘁

✅ Weekly newsletter: techstories.co/updates
✅ WhatsApp Channel (If 1-click doesn't work): techstories.co/wa

𝗣𝗵𝗼𝘁𝗼 𝗖𝗿𝗲𝗱𝗶𝘁: Unsplash/CHUTTERSNAP

Using Claude Code, a man accidentally broke into thousands of live DJI Romo robot vacuums, accessing camera feeds and remotely controlling them.

This #UnfilteredFirday, let's talk about IoT security. And where AI could get us.

✅ Join my WhatsApp Channel (1-click): https://lnkd.in/gfZUWPSe

𝟭/ 𝗪𝗵𝗼𝗼𝗽𝘀, 𝗜 𝗵𝗮𝗰𝗸𝗲𝗱 𝘆𝗼𝘂𝗿 𝘃𝗮𝗰𝘂𝘂𝗺

Hoping to link his brand-new DJI Romo vacuum to a PS5 controller for a geeky side project, a man fired up Claude Code and got to work.

By reverse engineering the protocol used to communicate with the vacuum, Sammy Azdoufal gained direct control of his robot vacuum.

But it turned out it also gave him access to thousands of other DJI robot vacuums too. Sammy could now:

- Check battery levels.
- View live camera feeds.
- Remotely control robots.
- Listen through microphones.
- Generate 2-D maps of each home.

What's more, with the IP addresses of the robots, he could also estimate the location of each device.

𝟮/ 𝗜𝗻 𝘀𝗶𝗺𝗽𝗹𝗲 𝘁𝗲𝗿𝗺𝘀

The DJI Romo uses the well-established and robust MQTT protocol to communicate with DJI servers and the official mobile phone app.

But while proper authentication was required to access a DJI Romo, poor design meant any authentication token could be used to control all connected robot vacuums.

This wasn't a problem with the official app, which restricted access to local devices. But when Sammy extracted his own authentication token* using Claude Code, it gave him access to every device out there.

*An authentication token is a temporary, unique code issued upon verification of a password, ensuring that the password doesn't need to be resent every time.

𝟯/ 𝗔𝗻𝘆𝘁𝗵𝗶𝗻𝗴 𝗰𝗮𝗻 𝗯𝗲 𝗵𝗮𝗰𝗸𝗲𝗱

When the story first broke earlier in the week on The Verge, DJI was already informed and it sent a statement that the issue was resolved prior to public disclosure.

Sammy claims other weaknesses remain, though he could no longer see or hear through other devices.

The issue here is weak app design on the part of DJI. However, it's worth noting how Claude Code lowered the barriers of entry, making it far easier to find or exploit security flaws.

The other consideration is this: If even a commercial product could contain such security oversights, what of vibe coded apps created by non-technical users?

I personally operate on the assumption that my home smart devices can be compromised. This is because cybersecurity is often last on the list for consumer products, and products are rarely - or never - patched after they are released.

Anyway, I'm curious about your thoughts on this: Are we facing a potential vibe-coded security apocalypse?

📆 𝗚𝗲𝘁 𝗺𝘆 𝗰𝗼𝗻𝘁𝗲𝗻𝘁

✅ Weekly newsletter: techstories.co/updates
✅ WhatsApp Channel (If 1-click doesn't work): techstories.co/wa

𝗣𝗵𝗼𝘁𝗼 𝗖𝗿𝗲𝗱𝗶𝘁: DJI.

AI is simultaneously creating and solving problems. Cybersecurity is just the start.

In my latest Substack piece, I look at why rapid AI-driven transitions are becoming the new normal, and why no industry will be spared.

You can read it here: go.clearlytech.co/fsOo

Here's the quick read version.

𝟭/ 𝗔𝗜 𝗮𝘀 𝗯𝗼𝘁𝗵 𝘁𝗵𝗲 𝗽𝗿𝗼𝗯𝗹𝗲𝗺 𝗮𝗻𝗱 𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻

(a) Looming vibe-coding security apocalypse

A man hacking his own robot vacuum found a flaw that gave him control over thousands of others. If that can happen with a commercial product, what about the flood of vibe-coded apps out there?

(b) AI as its own antidote

Anthropic's latest AI model has already uncovered over 500 vulnerabilities in mature open-source codebases - bugs that slipped past expert review for decades.

𝟮/ 𝗕𝗿𝗮𝗰𝗲 𝗳𝗼𝗿 𝗽𝗿𝗼𝘁𝗿𝗮𝗰𝘁𝗲𝗱 𝘂𝗽𝗵𝗲𝗮𝘃𝗮𝗹𝘀

My hypothesis: we can expect years of transitions even if AI stops getting better. Spoiler: there's no sign of AI slowing down. Every industry will eventually feel this.

tldr; We're far from understanding what AI means for cybersecurity.

How do you think AI will impact your industry?

📆 𝗚𝗲𝘁 𝗺𝘆 𝗰𝗼𝗻𝘁𝗲𝗻𝘁

✅ Weekly newsletter: techstories.co/updates
✅ WhatsApp Channel: techstories.co/wa

𝗣𝗵𝗼𝘁𝗼 𝗖𝗿𝗲𝗱𝗶𝘁: Unsplash/Jerry Wang

As the Head of the Department of Mechanical Engineering at the National University of Singapore (NUS), I lead strategic initiatives that position NUS ME as a global leader in sustainable energy solutions, robotics and machine intelligence, advanced manufacturing and materials. With a proven track record in securing multi-million dollar research funding and managing high-impact collaborations with industry leaders like Meta, Schneider Electric and Equinix, I specialize in translating cutting-edge research into actionable technologies.

Key achievements include:
Leading the Sustainable Tropical Data Centre Testbed (STDCT) to pioneer AI-driven cooling solutions for data centers, securing over S$25 million in research funding.
Co-authoring research on AI-driven data center cooling and sustainable energy systems, featured in top-tier journals like Applied Energy and Energy Conversion and Management.
Delivering keynote speeches at international conferences, covering sustainable AI infrastructure, advanced cooling, and energy transition.
I am passionate about leveraging my expertise to contribute meaningfully to sectors such as digital infrastructure, renewable/sustainable energy, robotics, AI and advanced manufacturing. I welcome opportunities to contribute strategic insights and governance expertise to organizations committed to innovation and sustainability.

Key Expertise:
Sustainable Cooling and Energy Systems
AI and Robotics for Industry 5.0
Strategic Industry Partnerships and Commercialization
Governance and Advisory for R&D Strategy

PS Lee

Geothermal AI Data Centres: Designing the AI Era from First Principles

AI is rapidly turning “data centres” into AI factories, continuous, high-density industrial plants where firm power, heat rejection, and water stewardship become the real constraints.

One of the most promising responses is geothermal-powered, geothermal-enabled AI data centre campuses, especially in geothermal-rich regions like Indonesia.

🔋 Beyond “Power + Cooling” as Separate Problems
Conventional design treats power and cooling as separate systems. That model breaks at AI scale.
Geothermal enables integrated thermo-electric design:
Firm 24/7 low-carbon baseload (critical for AI uptime and carbon credibility)
A thermal platform to support water-lean heat rejection (e.g., geoexchange / subsurface buffering)
Potential to reduce electric parasitics by shifting some cooling burden away from compressor-heavy chilling
The shift is from energy consumption to energy orchestration.

🌡️ PUE ≤ 1.15 in the Tropics—Hard, but Feasible
A credible target for AI halls is PUE ≤ 1.15 (annualised, operating)—with clear boundaries and auditable measurement.
To get there in hot-humid climates:
High liquid cooling coverage (direct-to-chip; immersion where useful)
Warmer facility water temperatures to minimise or avoid chillers
Geothermal-assisted / water-lean heat rejection (avoid over-reliance on evaporative towers)
Tight controls and continuous performance verification under worst wet-bulb weeks, not cherry-picked months

📈 Scale with Discipline: Phased Capacity that Matches Reality
At AI scale, the biggest mistake is “capacity vanity.” The right phasing is validation-led:
Phase 1 (10–15 MW IT): prove operating PUE/WUE, reliability, M&V
Phase 2 (30–60 MW IT): replicate in 10–15 MW blocks, refine integration and controls
Phase 3 (150–300 MW IT planned): AI factory scale, with an initial 60–120 MW build-out
This preserves learning loops while keeping the end state ambitious.

🌍 Why Indonesia Matters
Indonesia holds a significant share of the world’s geothermal potential and sits close to fast-growing digital economies. That creates a strategic opening: geothermal regions can become the next frontier for sustainable AI infrastructure, similar to how Nordic hubs leveraged climate and hydro.

⚙️ The Real Challenge: Systems Integration
The hardest part isn’t a single technology—it’s integration:
power quality and reliability for AI ramps
heat-sink sizing without long-term saturation
connectivity redundancy and bankable SLAs
transparent, audit-ready sustainability metrics
This is where thermofluids, power systems, controls, and digital twins must converge.
If done well, geothermal AI campuses can deliver globally competitive PUE in tropical climates, lower water intensity, and a credible pathway to low-carbon compute at scale.

#AIInfrastructure #DataCentres #GeothermalEnergy #LiquidCooling #Sustainability #EnergyTransition #GreenDataCenters #ThermalEngineering #ClimateTech #DigitalInfrastructure

Head of NUS Mechanical Engineering & Executive Director of ESI | Expert in Sustainable AI Data Center Cooling | Keynote Speaker and Board Member

No alternative text description for this image

taplio

The Battle

Paul Mah

PS Lee

Followers

Posts

Engagements

Likes

Performance ScorePaul Mah's LinkedIn Performance

Performance ScorePS Lee's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Paul Mah

PS Lee

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Paul Mah's LinkedIn Performance

Performance ScorePS Lee's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScorePaul Mah's LinkedIn Performance