Intel lost two senior AI executives within days: Chief AI Officer Sachin Katti departed for OpenAI on November 10, 2025, while datacenter GPU VP Saurabh Kulkarni joined AMD. The departures expose Intel’s struggle to retain talent amid failed Gaudi chip ambitions and Nvidia’s 94% AI market dominance. CEO Lip-Bu Tan now personally oversees Intel’s AI division.
Intel’s artificial intelligence division is hemorrhaging leadership at the worst possible time. The chipmaker faces a critical talent crisis after losing two senior executives to direct competitors within a single week, a stunning blow that raises serious questions about Intel’s ability to compete in the $165 billion AI chip market.
Table of Contents
What Happened at Intel
Sachin Katti Joins OpenAI’s AGI Mission
On November 10, 2025, Sachin Katti announced his departure from Intel after four years with the company, where he served as Chief Technology and AI Officer. Katti is joining OpenAI to design and build compute infrastructure for the company’s artificial general intelligence (AGI) research, a role that puts him at the center of one of tech’s most ambitious moonshots.
OpenAI President Greg Brockman called the hire a major win, stating he’s “incredibly excited” to work with Katti on infrastructure that will “power our AGI research and scale its applications to benefit everyone.” For Intel, it’s a devastating loss. Katti had led the company’s entire AI strategy since CEO Lip-Bu Tan reorganized management in January 2025.
Saurabh Kulkarni Moves to AMD
Just days before Katti’s announcement, Saurabh Kulkarni revealed on LinkedIn that he was leaving Intel for AMD. Kulkarni served as Intel’s Corporate Vice President of Datacenter GPU Product Management and played a key role in the company’s troubled Gaudi AI accelerator program.
The timing couldn’t be worse for Intel. AMD CEO Lisa Su announced the company is “on track” to generate tens of billions in annual revenue from its Instinct GPU business by 2027, a direct challenge to both Nvidia and Intel. Kulkarni’s insider knowledge of Intel’s AI roadmap now sits with a chief competitor.
Timeline of Departures
Intel’s AI leadership exodus didn’t start this month. According to CRN’s exclusive reporting, technical leaders Ronak Singhal and Rob Bruckner also departed Intel earlier in 2025 as part of a broader talent drain. CEO Lip-Bu Tan has been forced to recruit externally, bringing in former Apple chip designers Jean-Didier Allegrucci and Shailendra Desai to fill gaps.
Who Are These Executives
Sachin Katti: From IIT Bombay to Stanford to Intel
Sachin Katti graduated from IIT Bombay before spending nearly 15 years as a Stanford University professor, where he earned recognition for pioneering work in wireless networking and edge-based machine learning. Intel recruited him four years ago specifically to lead its AI transformation.
Katti’s academic pedigree and technical expertise made him one of Intel’s most valuable assets. His research background in distributed computing and network infrastructure directly aligns with OpenAI’s need to build massive-scale systems capable of training frontier AI models.
Saurabh Kulkarni’s Datacenter GPU Expertise
Kulkarni brought deep datacenter experience to Intel, having previously worked at Microsoft as director of engineering for cloud and AI system technologies. Before joining Intel in 2023, he served as chief product officer at Lucata and general manager at Graphcore, an AI chip startup that was acquired by SoftBank in 2024 after failing to compete with Nvidia.
His LinkedIn profile reveals he led “cross-functional product management across systems, software and silicon design” and championed Intel’s “rack-scale systems initiative” for datacenter AI. That expertise now belongs to AMD.
Why Intel Is Losing AI Talent
The Gaudi Chip Failure
Intel’s Gaudi AI accelerators were supposed to be the company’s answer to Nvidia’s datacenter dominance. Instead, they became a cautionary tale. Intel recently abandoned its goal of generating $500 million in revenue from Gaudi chips in 2024 after the products failed to gain significant customer adoption.
The company claims Gaudi 3 offers 1.5x faster training performance than Nvidia’s H100 on certain workloads and costs significantly less. But impressive benchmark claims mean nothing without customers. Intel’s inability to translate technical specs into market traction has become a pattern.
Nvidia’s Crushing Market Dominance
Nvidia controls an estimated 86-94% of the AI chip market, depending on the segment measured. Intel’s share is in the single digits. AMD’s Instinct GPUs are gaining ground with major cloud customers including OpenAI itself, which gives AMD credibility Intel lacks.
For ambitious AI executives, joining a company with less than 10% market share in the industry’s hottest segment isn’t appealing especially when competitors offer more resources, larger teams, and proven products.
Compensation Gap in AI Talent War
The AI industry is in an unprecedented talent war. Machine learning engineers command average salaries of $175,000, but top researchers and executives receive compensation packages worth tens or even hundreds of millions of dollars. Meta CEO Mark Zuckerberg has reportedly offered signing bonuses exceeding $100 million to lure talent from rivals.
Intel, which laid off 15% of its core workforce in 2025 as part of CEO Tan’s restructuring plan, simply can’t compete financially with cash-rich AI labs like OpenAI or GPU giants like Nvidia. When you’re cutting costs while competitors are printing money, the talent equation doesn’t work in your favor.
Intel vs AMD vs Nvidia: The AI Chip Battle
| Metric | Intel | AMD | Nvidia |
|---|---|---|---|
| AI Chip Market Share (2025) | ~8.7% | Growing (targeting tens of billions by 2027) | 86-94% |
| Flagship AI Product | Gaudi 3 (128GB) | Instinct MI300 series | H100, H200 |
| 2024 AI Revenue Target | $500M (missed) | $5.6B projected 2025 | $49B projected 2025 |
| Major AI Customers | Limited adoption | OpenAI, cloud providers | Dominant across all segments |
| Training Performance Claim | 1.5x faster than H100 (select workloads) | Competitive with Nvidia | Industry standard |
| Strategy | Open systems, cost advantage | Cloud partnerships, enterprise focus | Ecosystem lock-in, software advantage |
What This Means for Intel
CEO Lip-Bu Tan Takes Direct Control
Following Katti’s departure, Intel announced that CEO Lip-Bu Tan will personally oversee the company’s AI and Advanced Technologies Groups. In a statement, Intel emphasized that “AI remains one of Intel’s highest strategic priorities, and we are focused on executing our technology and product roadmap across emerging AI workloads.”
Direct CEO oversight can signal urgency and strategic importance, but it also reveals a vacuum. Tan is already managing an aggressive turnaround strategy that includes cost-cutting, asset sales, securing government CHIPS Act funding, and competing with TSMC in contract manufacturing. Adding direct AI oversight stretches leadership thin.
Intel’s New AI Strategy
At the 2025 OCP Global Summit in October, Intel unveiled a 160GB energy-efficient datacenter GPU as part of a new annual GPU release cadence. The company is pivoting toward open systems architecture and software platforms, a strategy designed to differentiate from Nvidia’s closed ecosystem.
Intel is also recruiting externally, bringing in former Apple chip designers to rebuild its AI silicon expertise. The question is whether external hires can replicate the institutional knowledge and customer relationships that walked out the door with Katti and Kulkarni.
Can Intel Catch Up?
Intel’s stock has gained approximately 92% in 2025, reflecting investor optimism about Tan’s turnaround strategy. But stock performance and product competitiveness are different battles. Intel continues to struggle with attracting major customers for both its AI chips and its contract manufacturing business.
The harsh reality: Intel is years behind Nvidia in AI software ecosystems, developer mindshare, and customer trust. AMD is outpacing Intel in GPU performance and cloud partnerships. Without transformative products or breakthrough customer wins, Intel’s AI ambitions may remain just that ambitions.
The Bigger Picture: Tech’s Brutal AI Talent War
Tech Industry AI Talent War
Intel’s losses are symptoms of a broader industry phenomenon. An analysis of OpenAI’s early GPT model research team shows fewer than half of the original authors still work at the company. Anthropic was founded by OpenAI defectors. Google DeepMind recently poached OpenAI’s Sora video AI lead Tim Brooks.
AI talent is the most valuable commodity in tech. Companies are competing not just with salaries, but with access to compute resources, cutting-edge research problems, and the opportunity to work on transformational technologies. Intel offers none of these advantages right now.
OpenAI’s Infrastructure Ambitions
Katti’s hire reveals OpenAI’s strategic priorities. The company isn’t just building AI models, it’s building the physical infrastructure to train artificial general intelligence at unprecedented scale. That requires expertise in distributed systems, networking, power management, and datacenter architecture, exactly Katti’s wheelhouse.
OpenAI’s compute needs are staggering. Training GPT-4’s successor will require thousands of GPUs (likely Nvidia H100s or H200s) coordinated across multiple datacenters. Katti will design the systems that make this possible.
AMD’s Growing Threat to Nvidia
AMD is executing a calculated strategy to chip away at Nvidia’s dominance. By winning OpenAI as a customer and targeting cost-conscious enterprises, AMD is positioning its Instinct GPUs as the “good enough” alternative that costs less and avoids vendor lock-in.
Kulkarni’s hire strengthens AMD’s datacenter credibility. He knows Intel’s product roadmap, technical weaknesses, and go-to-market strategy intimately. That intelligence is invaluable as AMD targets enterprise customers who might have considered Intel’s Gaudi chips.
Frequently Asked Questions (FAQs)
Why did Sachin Katti leave Intel for OpenAI?
Katti joined OpenAI to design compute infrastructure for artificial general intelligence research. OpenAI offers him the opportunity to work on one of tech’s most ambitious technical challenges with virtually unlimited resources, something Intel’s struggling AI division couldn’t match.
How much do AI executives earn at companies like OpenAI?
While specific compensation wasn’t disclosed, top AI researchers and executives command packages worth tens to hundreds of millions of dollars. Meta has offered signing bonuses exceeding $100 million for elite talent, and OpenAI likely matched or exceeded Intel’s offers.
What is Intel doing to replace these executives?
CEO Lip-Bu Tan is personally overseeing Intel’s AI division while the company recruits externally. Intel hired former Apple chip designers Jean-Didier Allegrucci and Shailendra Desai for AI leadership roles, though they lack the AI-specific expertise of departing executives.
Why did Intel’s Gaudi AI chips fail?
Despite technical performance claims (1.5x faster than Nvidia H100 on select workloads), Gaudi chips failed to gain customer adoption due to Nvidia’s software ecosystem advantage, limited developer support, and Intel’s reputation challenges in the AI market. Intel missed its modest $500 million revenue target.
How does Intel’s AI market share compare to Nvidia and AMD?
Nvidia dominates with 86-94% of the AI chip market. Intel holds approximately 8.7% of the AI training accelerator segment, while AMD is rapidly growing and targeting tens of billions in annual Instinct GPU revenue by 2027.
Can Intel still compete in the AI chip market?
Intel faces steep challenges but isn’t out of the race. The company is pursuing an open systems strategy, annual GPU releases, and competitive pricing. However, without breakthrough products, major customer wins, or a dramatically improved software ecosystem, closing the gap with Nvidia and AMD will be extremely difficult.
Featured Snippet Boxes
What happened to Intel’s AI executives?
Intel lost two senior AI leaders in November 2025: Chief AI Officer Sachin Katti joined OpenAI to build AGI compute infrastructure, while datacenter GPU VP Saurabh Kulkarni moved to AMD. CEO Lip-Bu Tan now directly oversees Intel’s struggling AI division.
Why are Intel executives leaving?
Intel’s AI executives are departing due to failed Gaudi chip adoption (missed $500M revenue target), Nvidia’s 94% market dominance, limited career growth prospects, and the AI industry’s talent war offering massive compensation packages often worth hundreds of millions of dollars.
Who is Sachin Katti?
Sachin Katti is an IIT Bombay graduate who spent 15 years as a Stanford professor specializing in wireless networking and machine learning. He served as Intel’s Chief Technology and AI Officer before joining OpenAI in November 2025 to design AGI infrastructure.
How much market share does Intel have in AI chips?
Intel holds approximately 8.7% of the AI training accelerator market in 2025, far behind Nvidia’s 86-94% dominance. Intel’s Gaudi chips failed to gain traction despite performance claims, while AMD’s Instinct GPUs are rapidly growing market share.
