DDN andBlackstone

The usual way to price storage

Storage is conventionally treated as a commodity. Capacity is measured in bytes, prices fall as flash and disk improve, and suppliers compete on cost per terabyte. On that view a $5 billion valuation for a private storage company needs an unusual explanation.

What DDN's customers are buying

A processor working on a large job needs the next portion of data before it can continue. If the storage system cannot deliver it fast enough, the processor waits. A hypothetical operator with a facility of ten thousand accelerators, each costing tens of thousands of dollars, is paying for a very large amount of capital that earns nothing while it waits. If a better storage system recovers five percent of that time, the gain is the output of five hundred accelerators, which is worth far more than the storage system costs.

The comparison customers make is therefore against idle time on a much larger investment, not against cheaper storage. Suppliers whose product governs the productivity of a customer's existing capital tend to be priced better than their category implies. The networking business described in the Mellanox note is a similar case.

The company and the terms

DDN's chief executive, Alex Bouzari, said that “Blackstone's support accelerates our mission to redefine the enterprise AI infrastructure category,” and Blackstone's Jas Khaira said that “we are thrilled to be the first institutional investor in DDN.” The release said the investment would “help fund DDN's continued rapid growth,” and it named BofA Securities as DDN's exclusive financial adviser (Pulse 2.0, reproducing the release). The release gave no revenue figure. Forbes has reported revenue near $400 million in 2024, about $500 million in 2025 and a projection near $1 billion for 2026 (AI Weekly). On those figures $5 billion is about ten times 2025 revenue and five times the 2026 projection.

What would have to be true

A valuation on a private company is an expectation and not a result. This one implies that enterprise demand for this kind of system grows substantially, and that DDN holds its position as larger storage vendors and the cloud providers address the same requirement. The revenue projection is the checkable part, and it had not been reached when this was written.

The strongest objection

The objection is that the valuation reflects the market for anything adjacent to AI spending. In that case it would fall with the sector, whatever DDN's storage does. The evidence for that view is the number of related businesses that raised capital on favorable terms over the same period.

The two readings predict different results. If the valuation tracks sector sentiment, DDN's revenue growth should look like everyone else's and its valuation should move with theirs. If it tracks utilization, DDN should keep winning in the largest installations, where idle processors cost the most, even if sentiment weakens.

What a reader should take from it

When a supplier of a component sounds like a commodity, ask what the customer would lose if the component underperformed. The answer sets the price the supplier can defend. The separate account of what the bank did in this transaction is in Deals From Palo Alto.