<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Vinod Bijlani</title><link>https://www.vinodbijlani.me</link>
<description>I design AI factories from model to megawatt, so enterprises and nations own their data, models and future.</description><item><title>AI factories, delivered: from factory floor to first token</title><link>https://www.vinodbijlani.me/blog/ai-factories-delivered</link><guid>https://www.vinodbijlani.me/blog/ai-factories-delivered</guid>
<pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate><description>Modular AI factories are arriving as pods: inside the pod, the types of modular offering, and the providers building them.</description></item><item><title>More than 100 GW planned in six markets. How much will be live by 2030?</title><link>https://www.vinodbijlani.me/blog/how-much-data-centre-capacity-will-be-live-by-2030</link><guid>https://www.vinodbijlani.me/blog/how-much-data-centre-capacity-will-be-live-by-2030</guid>
<pubDate>Mon, 09 Nov 2026 00:00:00 +0000</pubDate><description>Pipelines are not delivery schedules. Grid delays, cancellations and the gap between announced and energised capacity are already showing.</description></item><item><title>From data centres to AI factories</title><link>https://www.vinodbijlani.me/blog/from-data-centres-to-ai-factories</link><guid>https://www.vinodbijlani.me/blog/from-data-centres-to-ai-factories</guid>
<pubDate>Mon, 26 Oct 2026 00:00:00 +0000</pubDate><description>From a few kilowatts per rack to around 120 kW. A similar patch of floor, a very different demand on the building.</description></item><item><title>I flew into Taiwan thinking about chips. I flew out thinking about tokens.</title><link>https://www.vinodbijlani.me/blog/where-ai-gets-hard</link><guid>https://www.vinodbijlani.me/blog/where-ai-gets-hard</guid>
<pubDate>Wed, 21 Oct 2026 00:00:00 +0000</pubDate><description>The industry is solving the compute problem. Enterprises are now facing the consumption problem, and that needs a governed token factory.</description></item><item><title>Can an enterprise operate, trust and afford agentic AI?</title><link>https://www.vinodbijlani.me/blog/operate-trust-afford-agentic-ai</link><guid>https://www.vinodbijlani.me/blog/operate-trust-afford-agentic-ai</guid>
<pubDate>Mon, 19 Oct 2026 00:00:00 +0000</pubDate><description>One question followed me from Taipei to Hong Kong. Four themes came up in almost every conversation.</description></item><item><title>How much does an AI answer really cost?</title><link>https://www.vinodbijlani.me/blog/tokenomics-and-datanomics</link><guid>https://www.vinodbijlani.me/blog/tokenomics-and-datanomics</guid>
<pubDate>Thu, 15 Oct 2026 00:00:00 +0000</pubDate><description>The token bill is only one part of the economics. Tokenomics plus datanomics shows more of the total cost of enterprise AI.</description></item><item><title>Where should this actually run? A practical guide to hybrid AI</title><link>https://www.vinodbijlani.me/blog/which-ai-workload-goes-where</link><guid>https://www.vinodbijlani.me/blog/which-ai-workload-goes-where</guid>
<pubDate>Fri, 09 Oct 2026 00:00:00 +0000</pubDate><description>Recurring inference changes the placement economics. Cost, control and performance should decide where each AI workload runs, not habit.</description></item><item><title>AI factory economics gets interesting below the waterline</title><link>https://www.vinodbijlani.me/blog/ai-factory-economics-below-the-waterline</link><guid>https://www.vinodbijlani.me/blog/ai-factory-economics-below-the-waterline</guid>
<pubDate>Thu, 01 Oct 2026 00:00:00 +0000</pubDate><description>A GPU is a capital decision. Power, cooling, networking, utilisation and staging compound over the life of the AI factory.</description></item><item><title>AMD Instinct in six years: from 32GB on a GPU to 31TB in a rack</title><link>https://www.vinodbijlani.me/blog/amd-instinct-gpu-to-rack</link><guid>https://www.vinodbijlani.me/blog/amd-instinct-gpu-to-rack</guid>
<pubDate>Tue, 29 Sep 2026 00:00:00 +0000</pubDate><description>AMD&#x27;s data centre GPUs have moved from a single accelerator to a rack-scale system. The system around the accelerator now matters as much as the chip.</description></item><item><title>In AI factories, sustainability and economics are now the same conversation</title><link>https://www.vinodbijlani.me/blog/sustainability-and-economics-same-conversation</link><guid>https://www.vinodbijlani.me/blog/sustainability-and-economics-same-conversation</guid>
<pubDate>Fri, 25 Sep 2026 00:00:00 +0000</pubDate><description>At 1 GW of IT load, the gap between an average and a highly efficient PUE is around 3,940 GWh and $315M of energy a year.</description></item><item><title>Every AI token has a hidden factory behind it</title><link>https://www.vinodbijlani.me/blog/every-ai-token-has-a-hidden-factory</link><guid>https://www.vinodbijlani.me/blog/every-ai-token-has-a-hidden-factory</guid>
<pubDate>Tue, 22 Sep 2026 00:00:00 +0000</pubDate><description>Price per million tokens tells you what an application consumes, not how efficiently the infrastructure produces it. That needs another lens: $/FLOP.</description></item><item><title>An AI decision that arrives too late is not intelligence</title><link>https://www.vinodbijlani.me/blog/edge-ai-bridge-to-physical-ai</link><guid>https://www.vinodbijlani.me/blog/edge-ai-bridge-to-physical-ai</guid>
<pubDate>Mon, 21 Sep 2026 00:00:00 +0000</pubDate><description>Edge AI is the bridge to physical AI. Scaling it safely needs a distributed AI factory: local execution, central control and evidence flowing back.</description></item><item><title>Nearly 250 GW by 2030: the geography of compute is being redrawn</title><link>https://www.vinodbijlani.me/blog/data-centre-capacity-2030</link><guid>https://www.vinodbijlani.me/blog/data-centre-capacity-2030</guid>
<pubDate>Sat, 19 Sep 2026 00:00:00 +0000</pubDate><description>Mapping data centre capacity across 29 markets to 2030. Capacity is the headline; power, economics and value per megawatt decide how much becomes productive AI compute.</description></item><item><title>Global data centre capacity in 2026, and who actually controls it</title><link>https://www.vinodbijlani.me/blog/global-data-centre-capacity-who-controls-it</link><guid>https://www.vinodbijlani.me/blog/global-data-centre-capacity-who-controls-it</guid>
<pubDate>Tue, 08 Sep 2026 00:00:00 +0000</pubDate><description>Operational capacity across 20 major markets, and the four kinds of operator behind it. The winners may be those who get the most value from every megawatt.</description></item><item><title>Australia&#x27;s data centre capacity is set to more than double by 2030</title><link>https://www.vinodbijlani.me/blog/australia-data-centre-capacity</link><guid>https://www.vinodbijlani.me/blog/australia-data-centre-capacity</guid>
<pubDate>Thu, 27 Aug 2026 00:00:00 +0000</pubDate><description>The bigger shift is what we build on top of it: regional AI factories designed around energy, and a complete sovereign AI supply chain.</description></item><item><title>Asia&#x27;s AI infrastructure buildout: $460B, 240 projects and a fivefold rise</title><link>https://www.vinodbijlani.me/blog/asia-ai-infrastructure-buildout</link><guid>https://www.vinodbijlani.me/blog/asia-ai-infrastructure-buildout</guid>
<pubDate>Tue, 18 Aug 2026 00:00:00 +0000</pubDate><description>Asia&#x27;s AI factory capacity could reach 71 to 84 GW by 2030. Power, economics, security and governance will decide how much of it creates value.</description></item><item><title>The AI factory is the foundation. Orchestration makes it a platform.</title><link>https://www.vinodbijlani.me/blog/ai-factories-need-orchestration</link><guid>https://www.vinodbijlani.me/blog/ai-factories-need-orchestration</guid>
<pubDate>Mon, 17 Aug 2026 00:00:00 +0000</pubDate><description>Takeaways from the SUSE Summit Sydney customer panel on what it takes to move enterprise AI from pilots into production.</description></item><item><title>HPE Discover 2026: from choosing a model to operating intelligence</title><link>https://www.vinodbijlani.me/blog/hpe-discover-2026-operating-intelligence</link><guid>https://www.vinodbijlani.me/blog/hpe-discover-2026-operating-intelligence</guid>
<pubDate>Fri, 19 Jun 2026 00:00:00 +0000</pubDate><description>AI at scale is no longer a model challenge. It is an infrastructure, operations and economics challenge.</description></item><item><title>What AI governance can learn from the data governance era</title><link>https://www.vinodbijlani.me/blog/what-ai-governance-can-learn-from-data-governance</link><guid>https://www.vinodbijlani.me/blog/what-ai-governance-can-learn-from-data-governance</guid>
<pubDate>Fri, 29 May 2026 00:00:00 +0000</pubDate><description>The organisations that built governance before scaling their data programmes won. History is repeating with AI.</description></item><item><title>Decoding NVIDIA&#x27;s GPU alphabet: from V to GB</title><link>https://www.vinodbijlani.me/blog/decoding-nvidia-gpu-alphabet</link><guid>https://www.vinodbijlani.me/blog/decoding-nvidia-gpu-alphabet</guid>
<pubDate>Thu, 18 Dec 2025 00:00:00 +0000</pubDate><description>What the V, A, H, B and GB in NVIDIA&#x27;s data centre GPUs really mean, and why the roadmap reads as an AI infrastructure strategy.</description></item></channel></rss>