How Oracle is de-risking its AI boom

Oracle is the best performer in the S&P 500 on Wednesday after releasing strong third-quarter results after the close the prior day.

The database giant’s ambitious plans for AI infrastructure growth has been giving investors some anxiety lately, as it takes on worrying debt levels with reliance upon one jumbo-sized customer: OpenAI. But management has hit on two big ways to alleviate some of investors’ concerns:

• More customer variety, and

• Expanding without hurting its balance sheet and cash flows too much via a strategy of bring-your-own GPUs.

Clayton Magouyrk, CEO and director, said:

“We have signed more than $29 billion of contracts since then across multiple customers using that new model. A combination of bring-your-own-hardware and upfront customer payments enables us to continue expanding without any negative cash flow from Oracle. Of course, this $29 billion was in addition to other deals we signed this quarter.”

“We expect this to be viewed positively by investors across the capital structure that have expressed concerns about medium-term CapEx and financing required to support Oracle’s aggressive growth profile,” wrote Deutsche Bank analyst Brad Zelnick.

We might expect that this a BYOH approach might weigh on Oracle’s top-line outlook. After all, the dinner bill is going to be a fair bit less (assuming no corkage fee) if you’re bringing a bottle of wine versus buying it at the restaurant. But that’s not the case, as demand for AI computing is white-hot. Oracle reported $553 billion in RPO (remaining performance obligations, or backlog), up 325% from last year. The company also raised its sales guidance to $90 billion for its next fiscal year from a previous outlook of around $84.4 billion.

“It’s worth noting that the better outlook was not attributed to any single AI enterprise deal, rather multiple contracts,” Bank of America analyst Brad Sills wrote. “This suggests Oracle is benefitting from ramping AI adoption in the broader enterprise segment.”

Oracle can minimize the amount of time it spends burning cash by deploying fresh compute capabilities with haste, enabling it to work through that ample and growing backlog.

Magouyrk added:

“We optimize our data center construction through standardized designs. Our supply chain has improved with more suppliers and deeper relationships. We have tripled our manufacturing sites and increased rack output by 4x, all in the last year. We have scaled our installation processes to enable multiple phases of delivery in parallel. Time from rack delivery to revenue has reduced by 60% in the past several months.”

The company’s guidance for $50 billion in capex this fiscal year, which ends this quarter, remained unchanged. That would entail a sequential drop-off in investment from over $18 billion in Q3 to less than $11 billion in Q4.

Overall, the company sees the data center bottleneck to be its only major obstacle, and is ready for the task. Magouyrk explained:

“And so as our business is going through this hyper growth phase, [AI infrastructure backlog is] the only drag on profitability. But thankfully, we’re getting — we’re very good in getting better at delivering that capacity. That capacity when we deliver it is all already contracted for at a very profitable rate. So when you combine those things together, we’re extremely confident in both the capacity that we delivered and the continuing to increase profitability of our AI business.”