US
1,100 MW Signed: Who Gets Paid While Microsoft's Indiana Plants Still Wait
NiSource on 16 Oct signed Microsoft to about 1,100 MW from 2028. AGP and GenCo are paid; the only named plant is a 600 MW battery still on paper.
Sean
Writes Second Order by TKN — a plain-English map of AI infrastructure and semiconductors for non-specialist investors. Focus: who gets paid, where money sticks, and which step is stuck. Not investment advice.
Explained
Who Buys AI GPUs? The Hyperscalers Nvidia Gets Paid By First
AI GPU buyers are the companies that pay for accelerators. Nvidia split $48.7B hyperscale and $40.3B ACIE in the July quarter. Who pays, and where cash sticks.
Wire
Samsung agrees to TSMC 3nm base dies for Nvidia's custom HBM
Nvidia has asked Samsung, SK hynix and Micron to make the base die of its custom HBM on TSMC's 3nm process, a Korean paper reported on 18 October, and Samsung has agreed. TSMC is paid per wafer whoever wins; its 3nm lines are already its biggest node.
AI Data Centers
What Does IREN Do? From Bitcoin Miner to AI Cloud Data Centers
IREN is an Australian-founded AI cloud that owns its grid connections, data centres and GPUs: about 5 GW of grid agreements, a $9.7 billion Microsoft contract at Childress (first of four 50 MW halls accepted in August 2026), and Bitcoin mining it aims to retire by end-2026.
Wire
Anthropic offers Blackstone 10% to fund $60B of Google TPUs
Anthropic is offering Blackstone funds a 10% coupon on $9 billion of junior debt in a $60 billion plan for about 1.5 GW of Google TPUs, the FT reported on 16 October. The chip debt got bigger, and the riskiest layer got dearer than June's 8.5%.
Catalyst
Microsoft drops 23,000-GPU Essex site as UK power slips to 2030s
Microsoft has reportedly walked away from renting 23,000 Nvidia GPUs at Nscale's Loughton site after grid power slipped to the 2030s, so in Britain the connection date, not the chips or the tenant, now decides whether an AI hall gets built.
Explained
What Is Sovereign AI? Who Pays for a Country’s Own AI Compute
Sovereign AI is AI computing power, data and models a country keeps inside its borders and under its own laws. Governments rarely run the GPUs; regional neoclouds and telcos do. Nvidia is paid when the chips ship. The operator is paid only when someone signs for the hours.
Taiwan
Up to US$4 Billion a Park: Who Gets Paid as Taiwan Sends Its AI Suppliers to Texas
TSMC and Foxconn can pay for their own US factories. The small firms that make their boards, fans and power parts often cannot. This week Taiwan said it will help build US parks for them, up to US$4 billion each, and make unlicensed AI server sales to China a crime.
Analysis
If Neocloud Debt Stays Near 1,000 bp, Where Does Cash Stick?
CoreWeave's default swaps hit about 1,038 bp while rated GPU loans priced near 6-7%. The price of neocloud money now splits by customer. Who gets paid first, who waits, three scenarios with the arithmetic shown.
Wire
Modal raises $750M at $15.75B as AI cloud run rate tops $1B
Modal raised $750 million at a $15.75 billion valuation on 14 October after its run rate passed $1 billion. It sells AI computing by the second on capacity from more than 20 clouds, in the same week the chip owners' debt got dearer.
Malaysia
18 Sen to Use the Wire: Who Gets Paid as Malaysia Prices Green Power for Data Centres
Malaysia's data centres can buy solar power straight from a developer, but it still travels on TNB's wires. On 13 October the regulator set that toll at 18 sen a unit, not the 14 sen the market heard in September. The grid is paid either way.
Explained
What Is PUE? Power Usage Effectiveness in AI Data Centers
PUE is a data center's total energy divided by its IT energy. The industry averaged 1.52 in 2026; Google ran 1.09 in 2025. With the grid connection fixed, PUE decides how many megawatts a hall can lease, and the tenant usually pays for the overhead.