Scoreboard! Since we last wrote:

The US-Iran ceasefire has held, but no meaningful progress towards permanent peace has been made.

April price inflation printed higher than expectations.

Oil continues to march higher.

Bond yields continue to march higher.

The Fed is now more likely to raise interest rates than they are to cut them.

Trump flew to Beijing to meet Xi, but nothing substantial was announced.

Yet, AI stocks have marched higher and carried the Nasdaq up 6% over the last two weeks, supported by strong earnings growth and forward guidance.

Cisco led the charge reporting after market close on Thursday, May 14th. The company announced they had already exceeded their full-year 2026 target of $5B in AI-infrastructure orders and raised their target to $9B.

AMD announced $10B in Q1 revenue, up nearly 60% y/y.

Two days later, Coreweave announced their revenue doubled y/y, and that they now have a $99B backlog of orders that they’re working to fulfill.

A small sampling of AI performance, since May 5th:

NVDA +14%
AMD +21%
MU +19%
NBIS +28%
CSCO +26%

These names continued to move aggressively higher despite a series of negative macro developments.

Should the macro loosen up - mostly predicated by an end to the US/Isreal vs Iran War - we’re staring at a euphoric risk-taking environment into the upcoming SpaceX IPO.

Should the macro continue to worsen - missiles start flying, ground invasion begins, Hormuz explosions - it’s unlikely these AI names continue to float unscathed.

But significant dips will be aggressively bought on headlines that hint at conflict resolution.

Applied Digital (APLD) is not a chip company. It is not a software company. It is not pretending to have a chatbot. It is a datacenter “neocloud”, and the #1 ranked stock on May’s 5x Upside Rankings.

APLD builds and operates high-density data center campuses for AI and high-performance computing workloads. Hyperscalers need absurd amounts of powered, cooled, AI-ready capacity, and APLD is trying to become one of the few public-market ways to own that bottleneck.

That is the whole setup.

The AI trade started with GPUs. Then it moved to networking. Then cooling. Then power. Now the market is slowly figuring out that none of this works if hyperscalers cannot get large blocks of energized power into purpose-built campuses fast enough.

That is where APLD lives.

The company’s core asset is not “AI exposure.” It is powered land plus long-term hyperscaler leases. APLD’s job is to turn power access into AI factories, finance those projects without nuking common shareholders, and then collect long-duration infrastructure-style cash flow.

That is also why the bull case is so intoxicating.

APLD says it has roughly 1 gigawatt(GW) of critical IT load under construction, with about 900 MW fully contracted and operational across its AI Factory campus portfolio.

For comparison’s sake, 1 gigawatt powers the city of San Francisco for over a year.

Management’s April investor presentation points to roughly $23B of anticipated aggregate rental revenue across Polaris Forge 1, Polaris Forge 2, and Delta Forge 1, its 3 main datacenter campuses.

For a company with a current market cap around $12B, that is the part that makes the brain start doing irresponsible math.

The campus breakdown is the story:

Polaris Forge 1 — Ellendale, North Dakota:
APLD’s flagship campus. Roughly 400 MW contracted, with the first 100 MW facility already energized and operating. CoreWeave is the anchor tenant under long-term leases.

Polaris Forge 2 — Harwood, North Dakota:
A 300 MW campus, with 200 MW contracted to a U.S.-based investment-grade hyperscaler.

Delta Forge 1 — Southern U.S.:
A newer 430 MW AI Factory campus, with a 300 MW lease signed with a U.S.-based high-investment-grade hyperscaler. The company says this lease represents roughly $7.5B of total contracted value over an estimated 15-year term, with initial operations expected around mid-2027.

The perfect APLD customer is obvious: a hyperscaler or AI cloud operator that needs massive, high-density capacity faster than traditional data center development can provide it. Think customers with three traits:

CoreWeave is the archetype. So are large cloud platforms like Amazon, AI infrastructure specialists, and hyperscalers like Meta that cannot wait five years for the normal utility/data-center pipeline to catch up.

The deal pipeline matters because this business scales in chunks, not drips. APLD does not need to sell another $10/month SaaS seat.

It needs to lease another 100–300 MW block, secure project financing, and execute construction. Each signed campus can change the forward revenue base materially.

That is the magic and the danger.

The scaling case looks something like this:

If APLD can convert its current contracted campuses into operating assets, the market may stop valuing it like a risky developer and start valuing it more like a scarce AI infrastructure landlord.

Management’s presentation points to expected site NOI margins in the mid-to-high 80% range across the major AI campuses.

Even if investors haircut those targets, stabilized AI campus economics could be very powerful if financing costs remain manageable.

APLD has to fund construction, manage debt, avoid brutal dilution, deliver on time, and keep tenant credit quality strong.

That is the underwriting question:

Can APLD turn announced megawatts into energized megawatts, and then into cash flow that common shareholders actually get to keep?

The latest quarter gave the bull case something real to point at.

Fiscal Q3 2026 revenue was $126.6M, up 139% year over year. HPC hosting contributed roughly $71M of revenue, including about $44.1M of base rent, as the first Polaris Forge 1 building operated for the full quarter. Adjusted EBITDA was $44.1M.

But the GAAP picture is still messy. The company also reported a net loss attributable to common stockholders of roughly $100.9M.

On the chart, APLD is extended, but not broken.

As of the May 15 close, APLD was around $42.56, versus a 52-week low near $6.73 and a recent high around $46.71. The stock is up roughly 35% over the last 20 trading days and more than 500% over the past year.

It is trading above its 20-day, 50-day, and 200-day moving averages, with the 20-day around $37.97, the 50-day around $31.26, and the 200-day around $27.87.

That is a strong trend, but not a sleepy entry. RSI is around 61, so it is not screaming overbought, but the stock has already had a massive run. The cleanest technical read is that bulls want to see APLD hold the high-$30s/low-$40s breakout zone.

A failure back below the 20-day would suggest the recent move needs digestion. A sustained push through the mid-to-high $40s would put the market back in price-discovery mode.

The short-interest angle adds fuel. Yahoo data shows short interest around 30% of float, which means positive lease, financing, or construction updates can move the stock violently.

So the clean investor framing is this:

APLD is not safe. It is not cheap in the traditional sense. It is not a “buy it and forget it” compounder yet.

It is a high-beta, execution-heavy, capital-intensive way to bet that AI’s real bottleneck is powered infrastructure — and that APLD can become one of the scarce landlords sitting on the right side of that bottleneck.

The 5x case is:

APLD proves Polaris Forge is repeatable, Delta Forge validates the multi-campus model, hyperscaler leases keep stacking, project debt stays available, site NOI margins hold up, and dilution does not eat the upside before common shareholders get paid.

That is a lot of ifs.

But if those ifs start turning into operating assets, the market may decide APLD is not a data-center story stock anymore.

It may decide APLD is one of the toll booths for the AI power buildout. A $100B toll booth, at that.

Editor’s Note: We just released our monthly update to our list of stocks most likely to 5X. Check it out here:

https://www.thebrrr.com/upside-ranks

Their ops team built inventory command centers and reorder dashboards through Viktor. Supply chain gets daily stockout alerts before they happen. Marketing tracks ROAS and runs content calendars. CS has CSAT scores and support tickets triaged and briefed every morning in Slack, before the first support call. No dashboard digging.

48 internal apps, built through conversation. No code. No developer queue. Command centers, inventory dashboards, sales trackers, reorder systems.

That's one company. Across the platform, teams have built 2,000+ apps the same way: message Viktor in Slack, describe what you need, get a working tool deployed. No code. No six-week dev queue.

Your team doesn't wait for a product roadmap. They message a colleague.

5,700+ teams. SOC 2 certified.

"It was almost instantly adopted by the bulk of my team." — Boris Wexler, CEO, Space Dinosaurs

Start free. $100 in credits →

Got feedback? Follow the writer on Twitter @frank_locascio and send a message.

The BRRR is meant for informational purposes only. It is not investment advice. Please consult with your investment, tax, or legal advisor before making any investment decisions.