Oracle: The $455 Billion Question
Oracle has transformed itself from a legacy database company into one of the most aggressive players in AI infrastructure. The story here isn't about whether they beat consensus—it's about whether their unprecedented backlog translates into real momentum.
What to listen for on the call:
- Commentary on their contract backlog—essentially revenue that's already locked in but hasn't been recognized yet—currently sitting at $455 billion after a staggering 359% year-over-year jump
- Updates on multi-cloud database partnerships with Google, Microsoft, and Amazon
- GPU capacity constraints and how quickly they're bringing new data centers online
- Any color on the OpenAI relationship and whether additional hyperscaler contracts are materializing
π’ The bull case rests on Oracle Cloud Infrastructure (OCI) continuing to win business from AI companies that need massive GPU clusters. Larry Ellison has positioned Oracle as the cost-effective alternative to AWS, Azure, and Google Cloud for training large language models.
π΄ The bears, however, point to a critical concern: roughly two-thirds of that $455 billion of contract backlog won't convert to revenue for years, creating a significant gap between today's capital expenditures and future cash flows.
Key signals to watch:
- Are existing customers like xAI and OpenAI expanding their commitments?
- How aggressive is management's tone on FY2026 guidance?
- Any mention of new multi-billion dollar contracts being finalized?
- Data center expansion timeline—are they accelerating or hitting bottlenecks?
Oracle's stock has pulled back from its September highs, suggesting some skepticism has crept in. This earnings report is either a validation moment or a reality check.
Broadcom: The Custom Silicon Empire
Broadcom enters Thursday's report having just formalized a massive 10-gigawatt custom chip deal with OpenAI—a partnership that signals the dawn of what some analysts are calling the "XPU era."
Broadcom now operates two distinct growth engines...
1. The company's custom AI accelerator business (XPUs), where it co-designs and manufactures specialized chips for Google, Meta, ByteDance, and now OpenAI.
2. VMware, which Broadcom has aggressively restructured since the $69 billion acquisition
What matters most this quarter:
- AI semiconductor revenue trajectory—management guided for $6.2 billion in Q4, representing 66% year-over-year growth
- VMware Cloud Foundation (VCF) adoption rates and customer retention commentary
- Updates on the Thor Ultra networking chip and Ethernet-based scale-up solutions
- Any hints about new hyperscaler XPU partnerships beyond the current big four
The VMware integration is the wildcard. European regulators are scrutinizing the acquisition, and a coalition of cloud providers has filed legal challenges over licensing practices.
While the financial results have been stellar—infrastructure software margins have expanded significantly—customer sentiment remains polarized.
π© Red flags to monitor:
- Any softening in AI revenue growth guidance for FY2026
- Commentary suggesting VMware customer churn is accelerating
- Signs that non-AI semiconductor recovery is taking longer than expected
- Supply chain concerns around advanced packaging
No comments:
Post a Comment