Sachin Katti, who led compute scaling at OpenAI, said on X on 14 October 2026 that he is leaving the company to “build something new in a few weeks”, after securing what he called “double-digit GWs of compute” with partners including NVIDIA, Broadcom, AMD, Oracle, Microsoft, Google, AWS and CoreWeave.
Bloomberg reported the same day that OpenAI has named former xAI executive Uday Ruddarraju to lead its infrastructure and compute businesses. Katti wrote that he will continue as an adviser after 15 November.
Business Pill · CONCENTRATION
A one-minute explainer of concentration: when a few big pieces make up most of a total. It teaches the general idea only and says nothing about any organisation in this story.
The key insight: As we read it, Katti’s own list is the story: by his account OpenAI’s compute came from eight named partners across chips and capacity, and the unit that assembled it now passes, Bloomberg reports, to a former xAI executive.
Katti’s post opens: “After an incredible year at OpenAI, I’m moving on to build something new in a few weeks.”
He wrote that the team “secured double-digit GWs of compute across Ohio, Portsmouth, Georgia, Trainium and other projects”, and that it “got GPT-Infra off the ground and going in just three months.”
He described his path as running from “Stanford professor to entrepreneur, Intel business leader and CTO”, and now scaling AI infrastructure at OpenAI. He closed: “I’ll continue as an advisor after November 15. More soon!”
The eight compute partners Sachin Katti named in his 14 October 2026 post, grouped by us: chip suppliers highlighted (NVIDIA, Broadcom, AMD); the other five sell cloud or data center capacity. Trainium, also in his list, is AWS’s AI chip.
What Was Posted on X
After an incredible year at OpenAI, I'm moving on to build something new in a few weeks.
We secured double-digit GWs of compute across Ohio, Portsmouth, Georgia, Trainium and other projects with amazing partners including NVIDIA, Broadcom, AMD, Oracle, Microsoft, Google, AWS andโฆ
Sachin Katti’s own post of 14 October 2026, the primary source for this piece: he says he is leaving OpenAI to build something new, lists the projects and the eight partners behind its ‘double-digit GWs’ of compute, and says he will continue as an adviser after 15 November.
Who Takes Over
Bloomberg reported that OpenAI has named Uday Ruddarraju, a former xAI executive, to lead its infrastructure and compute businesses, “marking the latest leadership shakeup for its teams focused on expanding data center capacity.”
According to Bloomberg, Ruddarraju joined OpenAI last year and was appointed chief technology officer for compute in July. It said Katti, who had been overseeing the infrastructure unit, “will transition out of his role over the next few weeks and plans to remain as an adviser after that.”
The Partner List, Layer by Layer
Katti named eight partners. By our grouping, three are chip suppliers (NVIDIA, Broadcom and AMD) and five sell cloud or data center capacity (Oracle, Microsoft, Google, AWS and CoreWeave). Trainium, which he listed as a project, is AWS’s own AI chip.
As we read it, a buyer at this scale spreads its supply: by Katti’s account the gigawatts came from three chip suppliers, AWS’s Trainium and five sellers of capacity, not from one vendor.
As we read it, the handover matters because the commitments outlast the person: Katti says he stays on as an adviser after 15 November, and Bloomberg says Ruddarraju has been chief technology officer for compute since July.
As we read it, the same Wednesday showed the financing side of that list: CNBC reported CoreWeave’s 2032 bonds yielding 14.8% intraday, and CoreWeave is one of the partners Katti names.
Sachin Katti, on X, 14 October 2026
“We secured double-digit GWs of compute across Ohio, Portsmouth, Georgia, Trainium and other projects”
Three Implications
OPENAI’S COMPUTE PARTNERS Katti named NVIDIA, Broadcom, AMD, Oracle, Microsoft, Google, AWS and CoreWeave; Bloomberg says Ruddarraju now leads the infrastructure and compute businesses they deal with.
RIVAL LABS Bloomberg describes the change as the latest leadership shakeup for OpenAI’s teams focused on expanding data center capacity.
INVESTORS TRACKING AI CAPEX Katti put the compute secured at double-digit gigawatts but gave no total and no split by partner, so the size of each commitment is not in his post.
The framework puts it this way: “The $1T capex cycle is being underwritten by a $1T concentrated demand commitment from two counterparties whose combined free cash flow is deeply negative.” It names the two as OpenAI and Anthropic.
As we read it, Katti’s post shows one of those counterparties from the inside: a single team assembling double-digit gigawatts from eight partners, now under new leadership.
Where Katti said OpenAI secured its compute: Ohio, Portsmouth, Georgia and Trainium, AWS’s AI chip, “and other projects”, per his post on X of 14 October 2026.
What Is Not Established
Katti did not say what he will build next, and we do not speculate. He gave no gigawatt total beyond “double-digit” and no split by partner or site.
In the material we read, OpenAI had not published its own statement on the change, and Ruddarraju had not posted about it; his appointment rests on Bloomberg’s report.
Business Engineer Framework
The AI Capex Map and the State of AI Hyperscalers
A Business Engineer framework on who underwrites the AI capex cycle and where the demand commitments concentrate.
Sachin Katti said on 14 October 2026 that he is leaving OpenAI after a year in which, by his account, it secured double-digit gigawatts of compute with eight partners; Bloomberg reported that former xAI executive Uday Ruddarraju now leads OpenAI’s infrastructure and compute businesses.
A note on sourcing. On 14 October 2026 we read Sachin Katti’s own post on X in full, and Bloomberg’s report of the same day on Uday Ruddarraju’s appointment as far as the page served it. Katti’s words are his; the appointment is Bloomberg’s reporting. The grouping of his partner list into chip and capacity suppliers is ours. The CoreWeave bond figure is CNBC’s intraday reporting, which we covered separately.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur, HackerNoon and Search Engine People. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
