So many companies are making their own data centers. Who will actually benefit the most from this ? Which specific company or companies.

I’m buying into the SMR space to support the data centers. There’s no other feasible way to support the power demand besides nuclear, and full scale plants are too large.

OKLO, SMR, RYCEY

People tend to forget the law of supply and demand… Come 2027/2028 when most data centers are up and running, these neo-cloud GPU providers will have to lower prices due to excess supply. We’ll eventually see margin compression and the ones heavily leveraged will likely be under immense pressure…

The companies best placed are the ones utilizing their clusters for their own tangible products/services like Tesla (their massive clusters will power their AV vehicles as they roll out), Google for the same reasons, likely Amazon, and others deploying edge products and also developing their own clusters.

The reason companies like APLD can beat earnings massively is because supply has not caught up to demand. The writing is on the wall for these type of companies unless they use financial leverage to expand past being a neocloud provider.

I am in the data center business, all of these collocation companies are building data centers for that same 15 customers

CAT

Some software/cloud security providers: PANW, CRWD, ZS

I like the hyperscalers too, I’ve specifically chosen MSFT and GOOG. Amazon is a great choice too IMO.

I like VRT for data center cooling and reducing electricity use.

Also there are some longshots in the SMR OKLO, rolls royce, westinghouse, small modular reactor nuclear space. I’m sure that what data centers are doing in actual reality to achieve their short term electricity demands is not aiming for nuclear longshots, but they’re actually taking a multimodal approach: solar, grid contracts with grid utility companies, wind, petroleum generators, geothermal if they can get it.

Those who make chips, supply energy, and cloud

Data center growth? Look at STX, WD, SNDK, MU. Source: I’m in the industry.

The companies that own the datacenters and service the software.

Amazon, Google, Microsoft

ZS, PANW, CRWD

The chip producers will taper off (if) demand tapers off.

Utilities I don’t have a lot of knowledge on but classically they have capex and anti monopoly constraints for growth. Not really worth the investment in my opinion.

People claiming that the “AI Hype” will die down or stop don’t really understand whats happening right now in the software space. Models are being built out across industries for a vast array of utilities. From electric grid demand prediction, fine tuned language processing, medical imaging, crash analysis for insurance claims.

This isn’t a thing that’s just going to stop. It’s slowly going to be EVERYWHERE optimizing corporate efficiency in a variety of ways. Replacing jobs in some sectors and creating jobs in others.