Texas company to build data center near Alexandria that will employ 200

Applied Digital announced a $3.6 billion AI data center campus, Delta Forge 1, in Rapides Parish that is expected to create 200 permanent jobs paying an average of $90,000 and about 1,000 construction jobs before opening in summer 2027. The initial buildout will total 300 MW of critical IT load across about 300 acres, with Cleco providing power and the company’s waterless cooling technology supporting high-density AI workloads. The project received Louisiana sales and use tax exemption eligibility and underscores a broader AI data center buildout in the state.

Analysis

APLD is the cleanest direct beneficiary, but the more important read-through is that Louisiana is effectively socializing the permitting, tax, and infrastructure friction for hyperscale AI capacity. That improves the probability of follow-on deployments and raises the option value of adjacent land, transmission, gas generation, and fiber assets; in other words, this is less a one-off real estate project than a regional platform buildout. The market usually underprices this second-order effect because the first headline is jobs, while the real monetization path is power availability and interconnection priority.

The bottleneck is no longer demand; it is time-to-power. A 300MW campus of this type creates a multi-quarter backlog for transformers, switchgear, substations, gas peakers, and electrical contractors, which should keep pricing firm for the infrastructure stack even if AI capex sentiment cools. That favors names with transmission, grid, and EPC exposure more than pure data-center developers, because the latter can announce capacity faster than they can actually energize it.

META is a secondary winner if the Louisiana cluster lowers political resistance to large AI campuses and normalizes utility-supported expansion in the Southeast. But it also increases scrutiny around power consumption, which could slow timelines elsewhere if regulators start asking for more local concessions. The contrarian risk is that the market extrapolates too much from announcements: job creation is immediate optics, but equity value creation depends on power delivery and utilization, and those are 18-36 month variables, not headline-day catalysts.

The setup is bullish for infrastructure suppliers with limited incremental capacity and more mixed for data-center operators if capex inflation and power queues compress returns on invested capital. If energy markets tighten locally, the real hidden winner could be utilities and grid-adjacent contractors, while the hidden loser is anyone selling undifferentiated megawatt promises without secured interconnects. The probability-weighted takeaway is that this is a positive signal for the entire AI infrastructure complex, but the best risk/reward sits one level down the stack.

Key Decisions for Investors

• Long APLD into pullbacks over the next 1-3 months; thesis is that Louisiana validates the company’s ability to win scarce power-backed AI capacity, but size modestly because execution risk is still high and energization is a 2027 story.
• Pair trade: long utility/grid infrastructure names vs short a basket of pure-play data-center promoters over 3-6 months; the market will reward firms that can actually deliver transformers, substations, and interconnects, not just announce MW.
• Add META on weakness over 6-12 months; this improves the policy template for large AI campuses and reduces regulatory novelty risk, but use defined risk because the stock already discounts a lot of AI capex success.
• Consider long exposure to electrical equipment and EPC beneficiaries for 6-18 months; the order book impact from hyperscale buildouts typically shows up with a lag, but once it does, margins tend to re-rate before revenue catches up.
• Avoid chasing the headline in utility-free data-center names after a big move; if power availability becomes the gating item, announcements without secured load can reverse sharply once timelines slip.