What OpenAI makes for data centers
subsidiary · unit of Broadcom
Ownership: Oracle
OpenAI is a unit of Broadcom. It appears on 3 parts of the AI data center — Hyperscaler custom ASICs, Labs and Capital & financing — in the roles of manufacturer, operator and service. Those parts sit in Layer 1 — Compute silicon & packaging, Layer 9 — Building shell, site & construction and Layer 11 — Software, tools & services touching the hardware (optional).
Exposure
2 parts across 3 layers
No disclosed data-center revenue share on file.
Units and subsidiaries: MGX
Parts, by layer
Layer 1 — Compute silicon & packaging
| Part | Role | Confidence | Competitors | Note |
|---|---|---|---|---|
| Hyperscaler custom ASICs | manufacturer | inferred | Alphabet (Google), Microsoft, Alchip, Amazon (AWS), Broadcom, Marvell | custom accelerator |
Layer 9 — Building shell, site & construction
Layer 11 — Software, tools & services touching the hardware (optional)
| Part | Role | Confidence | Competitors | Note |
|---|---|---|---|---|
| Capital & financing (via MGX) | service | inferred | Apollo, Ares, Bank of America, BlackRock, Blackstone, Blue Owl |
Where OpenAI sits in the stack
OpenAI's placements in this catalog cover 3 parts, from layer 1 (compute silicon & packaging) to layer 11 (software, tools & services touching the hardware (optional)). None of them currently carries a sourced supply constraint tag. Across these parts it overlaps with other companies most often on the lines below.
What these parts are
- Hyperscaler custom ASICs — AI accelerator chips (often called XPUs) designed in-house by a hyperscaler for its own fleet, typically co-developed with an outside ASIC design partner and fabbed by a merchant foundry.
- Capital & financing — The private equity firms, credit funds, banks, and specialist lenders that finance data center construction and equipment purchases, provided here as context rather than as a hardware part.
- Labs — AI model developers whose compute needs — met through owned, leased, or partner-financed data center capacity — sit at the top of demand for the entire taxonomy below this layer.
Closest peers in this catalog
Latest feed
- OpenAI hires new energy regulatory lead — Data Center Dynamics, 2026-10-22OpenAI has hired Will Friedman from AWS to lead its energy regulatory efforts.
- OpenAI claims 100-plus secret files followed ex-employee back to Apple — The Register, 2026-10-15OpenAI alleges that a former employee transferred confidential hardware plans from Apple to OpenAI.
- OpenAI again shakes up data center leadership, with xAI exec Uday Ruddarraju to lead and Sachin Katti to leave — Data Center Dynamics, 2026-10-14OpenAI has announced a leadership change for its Stargate data center project, appointing Uday Ruddarraju from xAI to lead the initiative as Sachin Katti departs.
- Cerebras CEO reveals the capacity shortfall behind OpenAI's turn to Nvidia GPUs — DigiTimes, 2026-10-14The CEO of Cerebras stated that OpenAI transitioned to using Nvidia GPUs due to a significant capacity shortfall in alternative hardware offerings.
- Google data center strategic sourcing exec heads to OpenAI — Data Center Dynamics, 2026-10-13Google's strategic sourcing executive is joining OpenAI to focus on high-speed interconnects and rack power infrastructure.
- xAI's Michael Roberson joins OpenAI infrastructure team as commissioning program manager — Data Center Dynamics, 2026-10-09Michael Roberson is joining OpenAI's infrastructure team to manage commissioning programs for their data center operations.
- Cerebras' stock slide stops after Sam Altman calls it 'close partner' — DigiTimes, 2026-10-06Sam Altman described Cerebras as a close partner following a period of stock price volatility.
- Synopsys reshapes EDA, IP business models as AI shifts revenue from seats to customer output — DigiTimes, 2026-10-03Synopsys is evolving its EDA and IP business model to align with AI-driven shifts from seat-based licensing to performance-based revenue models.
- OpenAI’s Jalapeño ASICs are deployed alongside AMD EPYC ‘Turin’ CPUs as hosts, not Nvidia's Vera — Tom's Hardware, 2026-10-02OpenAI is deploying its custom Jalapeño ASICs using AMD EPYC 'Turin' CPUs as host processors instead of Nvidia's Vera platform.
- ‘This is how AI should be used’ — OpenAI head of hardware breaks down the AI-assisted design of its Jalapeño ASIC — Tom's Hardware, 2026-09-30The head of hardware at OpenAI discussed using AI-assisted design processes to develop the company's custom Jalapeño ASIC.
Questions about OpenAI
- What does OpenAI make for AI data centers?
- This catalog places OpenAI on 3 parts: Hyperscaler custom ASICs (manufacturer), Labs (operator) and Capital & financing (service).
- Is OpenAI publicly traded?
- Not on its own — this catalog records OpenAI as a unit of Broadcom, so the traded entity is the parent.
- Who else makes hyperscaler custom ASICs for AI data centers?
- Alphabet (Google), Microsoft, Alchip, Amazon (AWS), Broadcom and Marvell are also placed on Hyperscaler custom ASICs in this catalog.
- How well sourced are these placements?
- Every part→company placement carries a confidence tag. For OpenAI the mix is 3 inferred. "disclosed" means the company or a filing says so; "reported" comes from trade press or a reference-design list; "inferred" is a lead from the taxonomy with no citation yet. Treat reported and inferred as leads, not facts.
Answers are generated from this catalog's own rows — see sources & method.