Chris D’Andrea

Building The Compute Supply Index. Founder 3D OOH + AI Creative.

[Chris D’Andrea @chrisdandrea]

We're live tracking the AI datacenter buildout happening across the US. Beginning with a focus on Texas ~150 active listings and growing.

Something key stood out to us: ERCOT's queue has exploded to 225+ GW, but only ~5,300 MW of datacenter load is actually operational today. The gap between whats announced and built is crazy.

[Chris D’Andrea @chrisdandrea]

@Web3Marmot https://t.co/APp1qcWxhH

[Chris D’Andrea @chrisdandrea]

@andrewchen I built Reference, a CGI/3D OOH studio for brands like Google, Coinbase, Amazon, McLaren,
Now building Clarke: the compute supply index for AI infrastructure — tracking where AI data center capacity exists, what’s coming, and what will actually deliver.

https://t.co/QHNEXhJglI

[Chris D’Andrea @chrisdandrea]

@falcone_design 🫡

[Chris D’Andrea @chrisdandrea]

@iruletheworldmo Literally I can’t get this thing to move fast enough. More compute!

[Chris D’Andrea @chrisdandrea]

Data center industry insights are pointing now to something very big happening:

The next generation of facilities are going to dwarf what exists today.

We’re talking multi 10GW+ facilities across the US.

That’s like powering 5000 of NASA’s super computers at the same time - or one third of New York State’s grid draw.

Compute demand isn’t slowing, and despite the state-wide moratoriums and cancellations happening at the moment we’re still going to see compute supply growth that is going to make your eyes water.

The plans are just moving to states and communities that want them.

What exists today is just the foundation - compute at scale will really take hold of the country before the end of the decade.

[Chris D’Andrea @chrisdandrea]

@hthieblot Intelligence on data center buildouts
https://t.co/QHNEXhJglI

[Chris D’Andrea @chrisdandrea]

@milesdeutscher Looks like Clarke is the perfect fit for the Industrial AI sector. We’re building an intelligence layer for the AI data center buildout. All facilities get intelligent phase tracking and projections - when we have enough in a region, we roll that up into The Compute Supply Index.

[Chris D’Andrea @chrisdandrea]

We just applied Clarke to @speedrun! Building the compute supply intelligence that the AI infrastructure industry desperately needs. Let's GO! https://t.co/93gcgAzsib

[Chris D’Andrea @chrisdandrea]

@Jason I’m in

[Shaun Maguire @shaunmmaguire]

Build all the data centers

[Chris D’Andrea @chrisdandrea]

@tzartaraporvala @speedrun @tzartaraporvala I sent you a message on LinkedIn. Looking forward to connecting
https://t.co/QHNEXhJglI

[Chris D’Andrea @chrisdandrea]

@tkexpress11 Extremely excited for this!

[Chris D’Andrea @chrisdandrea]

@kenanhsaleh @a16z @speedrun Id love to chat - we’re building the index for compute supply: https://t.co/93gcgAzsib

[Chris D’Andrea @chrisdandrea]

@sickdotdev The compute supply index - we’re tracking every AI data center in the US and building projections of compute availability. https://t.co/QHNEXhJglI - taking alpha applications.

[Erin Price-Wright @espricewright]

Need a positive narrative for AI? We're not building a moon base, let alone colonizing Mars, without it.

[Chris D’Andrea @chrisdandrea]

@isareksopuro @mollytaft Same here, we are researching the same information for https://t.co/93gcgAzsib - it’s not just the utilities that won’t disclose, but orgs like Google cite their power numbers as proprietary. Even if this survey is done it may be years if the orgs provide full disclosures.

[Chris D’Andrea @chrisdandrea]

@duncancampbell The oklo data center designed by Gensler is in progress now. Also is using an SMR

https://t.co/Hdv2Mo4A20

[Chris D’Andrea @chrisdandrea]

Nice, check out what we’re working on at Clarke, lots of overlaps here. We’re tracking the AI buildout in real time through public documentation and permit data. Regional moratoriums like this are popping up in our system. and we provide estimates on when the projects will be completed.

[Chris D’Andrea @chrisdandrea]

We’re building a product that tracks the compute supply across the US - regardless if it’s hyperscale, or colo. Clarke tracks live MW phases and compute supply estimates based on publicly available documentation.

Over time, where compute is available and how much of it is coming on-line will become one of the largest growth factors in the AI agent economy.

[Mustafa Suleyman @mustafasuleyman]

For the next couple years at least, the entire AI industry is going to be defined by this fact: demand is going to wildly outstrip supply, and so what matters is which companies / products have margin to pay for tokens.

Those products will then rapidly improve because latency drives retention, and retention creates data to spin flywheels that improve the product and drive more adoption.

[Chris D’Andrea @chrisdandrea]

The first Clarke Alpha build is underway with layer 1 focusing on Texas/ERCOT. All core data is sourced from permits with compute estimates and construction timeline projections.

First publicly available version coming soon with 60us AI data centers records live and over 1200 pending. Full deep intelligence profiles fueling the future compute capacity index. Regional and custom reports will be available for enterprise customers starting Day 1.

If you're in the industry and would like to help us as a design partner we want to hear from you! Now taking Alpha applications: https://t.co/Aau9uHCAfW