AI hardware is bigger than Nvidia and the hyperscalers
Investors looking for the generative-AI buildout can widen the lens beyond the obvious winners and hunt for the less crowded infrastructure plays.
17 briefings · “Infrastructure”
Investors looking for the generative-AI buildout can widen the lens beyond the obvious winners and hunt for the less crowded infrastructure plays.

Upwind's new "Security for AI" push reframes AI risk as infrastructure, not a side product, and that changes how buyers will budget, govern, and deploy.

A former IBM cybersecurity chief says the companies concealed repeated intrusions to keep federal business, raising the stakes for contractors and regulators.
As AI infrastructure bills explode, Meta is testing a cheaper, faster build method that could reshape how big tech adds capacity.

The telecom operator is pushing deeper into digital infrastructure by pairing capital with Shorooq’s startup muscle, a move that could reshape how regional founders access scale.

The UAE deal pairs sovereign AI infrastructure with dialect-aware voice automation, aiming to move enterprises and governments from pilots to production across the GCC.

Brett Winton is anchoring ARK's SpaceX thesis on Starlink, a signal that satellite broadband is now being treated like core infrastructure, not moonshot fluff.

A security-focused rsync update broke some incremental backups, then commit history showing Claude-assisted work turned a bug report into a broader fight over AI in critical infrastructure.

Intel is betting a cheaper, memory-heavy datacenter GPU can win the prefill work Nvidia paused, which could reshape AI infrastructure buying decisions.

The London AI video company is tightening a screen-at-creation system that could shape how fast AI-generated media gets governed, approved, and shipped.

India, Brazil, the UAE, and Africa are building local AI stacks to work around scarce compute, changing where infrastructure power sits.

The revolt puts a sharp question in front of Amazon's leadership and peers: how do you justify mass cuts while funding a historic AI buildout?