This website uses cookies
Read our Privacy policy and Terms of use for more information.
OpenAI's Luna hit $0.20 per million tokens; Anthropic's Opus 5 launched at $5, half its predecessor's price.
Daily
But don't call it vendor financing.
Apollo
+27
Nevada, Mississippi, and Wisconsin all moving against hyperscalers at once. The grid is fighting back.
ERCOT's queue could hit 120GW of peak demand by 2030, but 80% of new projects have no power lined up.
80+ hardware vendors are already building to the new spec, with MGX racks shipping H2 2026.
Domestic accelerators, no US chips. The national compute grid is taking shape faster than Washington expected.
Or maybe it's a marketing ploy. Who knows?
70,000 NVIDIA GPUs by 2027, part of a 2GW regional buildout. Private compute is outpacing national grids.
The largest new power plant in the US will burn gas to feed data centres, and its builder has net-zero pledges.
At a $3.3bn valuation. Arm and Netflix's co-founder are in. First chip: 2027.
DeepSeek
One gigawatt, one campus: the largest GPU infrastructure lease on record squeezes every neocloud in between.
+1
OpenAI and Anthropic's margins are collapsing as Chinese rivals race to the bottom on inference pricing.
Demand is growing 3.6% a year through 2030, but 56% of US utility staff have under 10 years on the job.
Because nothing accelerates regulatory capture quite like fear.
Transmission takes a decade to build. Data centres need it in months. ESIG just mapped the gap.
The hyperscaler premium that defined GPU scarcity is evaporating as B200s flood the market.
Because who needs permits for a $2.8B turbine spend over three years.
Sale and leaseback becomes the model.
SK Group
+25
Nadella pitched in-house models as 40% more efficient than OpenAI and Anthropic, backed by $41B in AI capex.
The DOE axed 321 grants over state politics. The line was 4-6 months from done, saving $200M.
CXMT's Shanghai debut hands China a war chest to chase the HBM capacity SK Hynix and Samsung control.
The SEA-H2X submarine cable is live across Asia-Pacific, and it's a direct play for infrastructure sovereignty.
CEO Clem Delangue says the agent roamed Hugging Face's systems undetected for multiple days.
The AI lab is serious about breaking free from NVIDIA's supply chain. Here's what it needs.
And they're running on two-generation-old chips.
