Tag
A Stanford research paper indicates that small language models running on local devices can achieve performance comparable to large language models in data centers, potentially impacting hyperscalers' infrastructure investments.
Utility companies like Constellation Energy, NextEra Energy, and Vistra are reporting concrete earnings growth and raising guidance due to AI data center power demands, backed by multi-year contracts with hyperscalers such as Meta, Microsoft, and Google, though risks exist if AI spending slows.
A new Noreva report warns that hyperscalers like Amazon, Google, Meta, and Microsoft may regret betting on natural gas to power AI data centers, as prices could triple in some U.S. regions due to surging demand and limited supply growth.
A tweet outlining a weekend guide for shorting Nebius (NBIS) by reading hyperscaler AI cloud earnings, and going long Lululemon (LULU) as a hedge.
Analysts estimate that more than 70% of Amazon, Microsoft, and Google's AI revenues come from OpenAI and Anthropic, raising questions about whether the AI demand boom is sustainable or a bubble.
The article examines the growing compute shortage in AI, highlighting simultaneous bottlenecks across GPUs, memory, TSMC capacity, and power infrastructure, and how this scarcity is reshaping corporate strategy and market leadership.
This article discusses growing concerns from banks and hyperscalers about a potential AI bubble, citing a Bank for International Settlements report warning of economic risk and Oracle's significant stock decline. It features a podcast episode analyzing the situation.
An analysis of the circular financing structures behind neocloud companies like CoreWeave and Nebius, which rely on Nvidia equity, hyperscaler contracts, and GPU-backed debt to fund massive AI infrastructure buildouts, raising questions about sustainability.
Analysis of AI infrastructure spending, with Sequoia's David Cahn calculating $1.5 trillion in spending for 2026 and a required $3 trillion in revenue to justify it, while Apollo's Torsten Slok warns of recession risk if hyperscalers fail to meet cash-flow goals.
The article analyzes US grid capacity constraints and models the shortfall that must be filled by behind-the-meter power solutions for datacenters, projecting over 40GW by 2028.
Amazon has a leading position in data center capacity and power for AI, but Google is expected to close the gap by 2030.
Discusses the cracking of nuclear regulatory barriers, driven by hyperscalers' demand for energy, and highlights key thinkers on the shift.
This report from Multiples.vc provides public AI valuation multiples as of June 2026, covering hyperscalers, semiconductor supply chain, neoclouds, and other segments with median forward revenue multiples and growth rates.
An analysis of declining token prices for AI models despite new releases like GLM 5.2 and Kimi 2.7, suggesting possible diminishing returns from expensive models.
An academic study presented at the Americas Conference on Information Systems maps five systemic tensions from AI's data-centre boom, including energy paradox, water strain, hyperscaler dominance, sovereignty erosion and urban displacement, highlighting the growing environmental and social costs.
Cisco is repositioning itself as a core AI infrastructure supplier, reporting strong hyperscaler orders and a broadening market beyond GPUs to networking, security, and observability. The company raised its FY26 hyperscaler AI infrastructure order expectation to ~$9B and highlighted Silicon One as a critical differentiator.
Michael Burry warns that the AI boom may be built on temporary demand from hyperscalers training models, creating a 'bullwhip effect' that could lead to severe oversupply and a sharp correction for Nvidia.
Gulf states' AI ambitions are threatened by their heavy reliance on a few vulnerable undersea cables, risking disruptions to data flows and economic transformation.
The article argues that AI is too expensive to be economically viable for most companies, with hyperscalers spending trillions on data centers but failing to generate proportionate AI revenue. It suggests only hardware suppliers like NVIDIA benefit from the current AI bubble.
The article argues that the high capital expenditure, power infrastructure, and GPU costs make AI development economically unsustainable for all but the largest hyperscalers like Google, Microsoft, Amazon, and Meta.