Put the announced money next to the announced concrete and the sizing explains itself. Al Moammar's expanded contract runs 250 megawatts for more than 8.76 billion riyals, about $2.34 billion [14], which is roughly $9.4 million a megawatt [1], and at that rate the six gigawatts HUMAIN says it will have in the kingdom by 2034 [9] implies a build cost near $56 billion [2], above the $42 billion the estimate puts on Saudi AI and cloud expansion by 2030 [16] (different end dates, so treat the pair as a range and not a contradiction). A $2.5 billion vehicle [2] is about $417,000 per planned megawatt [3], under 5% of the per-megawatt cost the Al Moammar contract implies [4] and roughly 6% of the 2030 requirement [5]. At that size, the fund reads less like construction capital and more like a mechanism for buying signatures on the load.
Amin's claim that computing in Saudi Arabia costs about 30% less than elsewhere [13] implies buyers outside pay something like 1.43 times the in-kingdom price [8], and a gap that wide ought to fill halls unassisted, without a placement requirement written into the equity [3]. That the requirement exists anyway is the most informative line in the interview. Either the realised discount is narrower than the headline, or, the version I find more interesting, HUMAIN needs contracted load specifically because the buildout is partly financed by borrowing against offtake agreements Amin calls bankable [11]; a lender underwriting an offtake is underwriting the tenant's credit, so owning equity in the tenant is collateral maintenance wearing a venture jacket. The energy ministry's 16-gigawatt supply commitment against a 6-gigawatt plan [10] leaves ten gigawatts of headroom [7], which is a great deal of empty room to fill by invitation.
The machinery around it is already committed to dates. A 100-megawatt first phase with DataVolt at Neom's Oxagon is slated for 2028 [15], offices are planned in the United States, Saudi Arabia, France and the United Kingdom to speed deal flow [5], and Amin says he sees no obstacle to obtaining the chip capacity he needs [18]. HUMAIN is owned by the Public Investment Fund and chaired by Crown Prince Mohammed bin Salman [21], which is why a portfolio company's compute placement is a state objective before it is a term sheet.
The precedent Amin leans on is February's $3 billion xAI stake, converted to SpaceX equity after the merger [7], and PIF's disclosed $26 billion SpaceX holding [8] is 8.7 times that check [6], though the source does not break out how much of the holding came from the conversion, so read the multiple as a ceiling on the boast rather than a return. This is probably wrong, but I read the placement condition as a selection filter that works against HUMAIN: founders with cheap unconditional capital will not encumber where their workloads run, so the checks drift toward those who need money more than they need Riyadh, and then the portfolio and the offtake collateral sour together. A named tenant with real alternatives signing anyway would falsify that. So would the larger vehicle Amin floated to Semafor, above $10 billion [1], which would make this fund a quarter-sized pilot at most [9]. The ticket sizes will say more than the totals.
