I’ve Sat Through 3 Data Center Power Meetings This Month—No One’s Talking About Plug Power

(SeaPRwire) –

By: Ethan Gallagher
Plug Power’s 8% Thursday rally is a masterclass in sector hype outrunning operational reality. I’ve sat through three data center power procurement meetings in the last month. Not a single one brought up Plug as a serious contender for AI workload power needs. The stock’s gain has nothing to do with Plug’s own operational breakthroughs. It’s a spillover effect from Bloom Energy’s blockbuster quarterly report. Most retail investors are buying the “fuel cell + AI” narrative, not the actual performance of the company behind the ticker. This kind of guilt-by-association rally never lasts. It only sets up retail shareholders for more pain when the hype fades. The gap between real AI power winners and also-rans is growing too wide to ignore.

The official narrative frames the day’s moves as a broad fuel cell sector rally, lifted by AI power demand. Plug gained 7.9% to around $2.05 during afternoon trading. The advance snapped a six-session losing streak for the company. It briefly pushed the stock above the psychologically important $2 level.
PLUG Stock Card
FuelCell Energy surged more than 28%. Ballard Power rose nearly 9%. The trigger was Bloom Energy’s second-quarter earnings release. Bloom posted record revenue of $1.065 billion, up more than 165% from a year earlier. It generated positive operating cash flow of $226.4 million. Its gross margin came in above 33%. Bloom’s management said the company is a preferred provider of onsite power for AI data centers. Mizuho upgraded the stock to “Outperform” and set a $242 price target. The bank cited Bloom’s speed-to-power advantage and substantial financing capacity. Onsite power for AI data centers is not a niche market. Hyperscale operators are racing to stand up new clusters as fast as possible. Grid capacity in many key markets is already maxed out. Fuel cells offer a way to deploy reliable power in months, not years. That speed is worth a premium for AI operators losing money every day their clusters are offline. The real subtext here is far less rosy for the broader sector. Bloom outperformed Plug by more than 17 percentage points on the same day. That gap is not a rounding error. It’s the market drawing a clear line between a company that delivers real AI power solutions, and everyone else. A cleantech equities colleague joked last week that any random fuel cell ticker would jump after Bloom’s print. He wasn’t wrong, but the size of the gain tells you who the market actually believes in.

The official Plug turnaround story paints the company as a work in progress, closing the gap with leaders. The stock had fallen from $2.27 on July 21 to $1.90 on July 29. That was a decline of more than 16% over six consecutive sessions. Thursday’s rebound recovered only about 40% of that drop. Shares still trade roughly 10% below their level from late July. The financial gap between Bloom and Plug is stark. Bloom’s quarterly revenue was roughly 6.5 times larger than Plug’s most recent quarterly revenue of $163.5 million. Bloom reports positive operating margins and strong cash generation. Plug continues to post negative margins and significant cash outflows. Plug has made some progress compared with last year. First-quarter revenue increased 22%. Gross margin improved substantially from deeply negative levels. The adjusted loss narrowed. Chief executive Jose Luis Crespo has said the company remains on track for positive EBITDAS in the fourth quarter. That target is non-GAAP and forward-looking. Investors are now focused on Plug’s liquidity, upcoming quarterly results, and potential data-center power contracts. Liquidity concerns have hung over the stock for months. The company’s cash burn rate remains high. A lack of large, high-margin contracts could force it to raise more capital soon. Any new capital raise at current price levels would dilute existing shareholders heavily. The subtext here is that Plug’s turnaround thesis rests on very thin ice. The 10% gap from late July levels shows investors are not fully buying the recovery story. Non-GAAP EBITDAS targets exclude a long list of real operating costs. Data center operators don’t make procurement decisions based on adjusted earnings metrics. They care about proven uptime, deployment speed, and access to financing. Mizuho specifically called out those exact strengths for Bloom. Plug has not demonstrated any of them at hyperscale AI data center sites. I’ve talked to two data center facility managers who say they won’t consider a fuel cell vendor without three years of 99.99% uptime data at scale. Plug does not have that track record yet.

The AI data center fuel cell supply chain will consolidate around two or three proven players with low-cost capital access. Plug will either land a major, publicly disclosed data center power contract in the next two quarters, or it will be locked out of the AI power boom entirely.

Author bio: Ethan Gallagher, a Silicon Valley hardware architect with 15 years of experience designing and scaling data center power infrastructure strategies.