Incident sheet · /infrastructure
- System
- Infrastructure & Cloud
- Filed
- Read
- 5 min · 902 words
- Severity
- Routine
Nvidia Puts Up $118B in Guarantees to Unlock AI Data Center Buildout
Nvidia posted $107.6B data center revenue and disclosed up to $118B in AI infrastructure guarantees, as its board ruled out Jensen Huang's return and supply commitments hit $390B.

Event log
- Nvidia reported $107.6B Q3 data center revenue, up 110% year over year, and up to $118B in AI infrastructure guarantees
- The board unanimously rejected reinstating CEO Jensen Huang, who resigned Nov. 12; CFO Colette Kress is interim CEO
- Supply commitments rose from $279B to $390B in one quarter, driven mostly by multiyear memory procurement
- The largest guarantee, up to $105B, backs SB Energy's 4.25 GW PORTS-Pike campus in Ohio leased to OpenAI for 20 years
- Nvidia guides Q4 revenue to $130B plus or minus 3%, assuming no data center compute revenue from China
Nvidia disclosed up to $118 billion in financial guarantees tied to AI infrastructure projects on Tuesday, alongside data center revenue of $107.6 billion — up 110% from a year earlier. The guarantees put the chipmaker on the hook for financing arrangements that help customers secure the land, power and facilities needed to deploy its systems before a single server ships.
The quarter's headline numbers landed amid a leadership crisis that has dominated headlines worldwide. CEO Jensen Huang resigned Nov. 12 after reports of multiple affairs, including one with a woman U.S. intelligence officials suspect of ties to Chinese intelligence. CFO Colette Kress is serving as interim CEO while the board searches for a permanent successor.
A campaign to bring Huang back drew prominent backers — Elon Musk, Epic Games CEO Tim Sweeney and President Donald Trump among them — and a public petition collected more than 180,000 signatures, thousands from self-identified Nvidia employees. Huang said Tuesday he would be honored to return if the board decided it served the company's interests. It won't. Chairman Harvey Jones told the earnings call that directors unanimously accepted the resignation and will consider internal and external candidates instead.
What does $118 billion actually guarantee?
For data center operators, the guarantees address a constraint that comes before deployment: financing and building facilities with enough power for large AI workloads.
The largest piece backs SB Energy's PORTS-Pike Technology Campus in Ohio — credit support for roughly 4.25 gigawatts across nine buildings, hosting Nvidia infrastructure under 20-year leases to OpenAI, subject to limited exceptions. Each guarantee takes effect only after its associated building is completed, with the first expected in fiscal 2029. Exposure will peak at $105 billion, then decline over the lease terms as OpenAI makes payments. Nvidia had made no guarantee payments as of Oct. 25.
The remainder includes $9.5 billion in power and systems credit support guarantees and $3.5 billion in land, power and shell guarantees for neocloud providers. Nvidia separately reported $32 billion in future commitments under neocloud agreements, which depend on providers securing financing, buying Nvidia infrastructure and operating the capacity.
Stephen Sopko, semiconductor practice lead at HyperFRAME Research, cautioned against reading the full figure as neocloud credit risk. "Of $118 billion in guarantees, $105 billion backs land, power and shell for one campus, PORTS-Pike, which is leased to OpenAI for 20 years," he said. "The neocloud commitments are $32 billion and only apply once the operator secures its own financing and the capacity is accepted."
Nvidia locks in supply years ahead
Supply commitments climbed to $390 billion at quarter's end, up from $279 billion three months earlier — a $111 billion jump the company attributes mostly to multiyear memory procurement. The rest reflects manufacturing deals including new substrate agreements with Kinsus and Samsung Electro-Mechanics, plus packaging expansions at Siliconware Precision Industries (SPIL) and Amkor Technology. Kinsus plans to boost advanced ABF substrate capacity dedicated to Nvidia by more than 3.5 times, targeted for Q3 2028; SPIL's expansion runs through mid-2029.
"It's no longer just leading-edge wafers. It's memory, substrates and packaging, and that capacity takes years to build," Sopko said. Nvidia also works with IBIDEN, King Yuan Electronics and Unimicron Technology, and its Vera Rubin platform will draw on the same TSMC 3-nanometer, 5-nanometer and CoWoS manufacturing base the broader industry depends on.
"Nvidia is locking in supply two to three generations ahead, which squeezes everyone else, including custom ASIC programs, for that same capacity," he said.
Nvidia also reported $102 billion in combined cloud service agreements and data center lease commitments, up from $74 billion the previous quarter — infrastructure for chip engineering, product testing and AI model development.
Vera Rubin ramps, power caps persist
Volume shipments of Vera Rubin NVL72 systems began during the quarter — what Nvidia calls the fastest product ramp in company history. Racks are live at CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius, with systems expected soon at SpaceXSI. Oracle announced general availability of NVL72 systems and Vera CPUs; Google made its Vera Rubin-powered A5X instances generally available. Meta plans to deploy Vera Rubin systems, Vera CPUs, Spectrum-X networking and BlueField-4 DPUs.
Nvidia said Vera Rubin delivered 10 times more tokens per watt than GB300 on the MLPerf Endpoints benchmark — a specific test result, not expected fleet-wide electricity savings. "In power-capped sites, yes. Watts are the scarce input, so more tokens from the same grid allocation is real revenue capacity. But that 10x is an MLPerf result versus GB300, not a fleet average," Sopko said. He expects operators to spend efficiency gains on bigger models and longer-running agents: "Cost per token falls; total power demand doesn't."
What does this change for buyers?
Total Q3 revenue reached $115.8 billion, up 103% year over year. Net income was $65.3 billion versus $31.9 billion a year ago; GAAP diluted EPS hit $2.70, up from $1.30. Nvidia guides to $130 billion in Q4 revenue, plus or minus 3%, assuming zero data center compute revenue from China, with gross margins of 71% to 72% on the Vera Rubin ramp.
"The bigger shift is that powered shells are now the bottleneck, and Nvidia is acting as a credit enhancer to get them built," Sopko said — selectively, as other capital sources mature.
via techtarget.com (Source)
More from Olivia Hart
Show full bio
News editor covering money, funding and market shifts around enterprise information technology and the leaders who run at CIO Monitor.
207 articles
The newsroom digest
One email a day with the stories that matter. No noise.
Cross-references
- Nvidia's board ousts Jensen Huang; Musk calls the move a mistake
- Jensen Huang resigns from Nvidia as spy-affair scandal reaches the board
- Altman: AI Industry Will Be 'Fine' Without Jensen Huang at Nvidia
- Nvidia’s first earnings test without founder Jensen Huang
- Huang Says He Would Return as Nvidia CEO — Board Isn't Budging