As the world's leading company by market capitalization, how does NVIDIA continue its upward trajectory? The answer is straightforward: recruit more talent, and recruit aggressively.
Throughout 2025, Jensen Huang has been expanding the management ranks while spending to acquire entire teams—a dual strategy of poaching key executives in markets, policy, and human resources, coupled with acquiring startups to "package-deal" their technical leadership into the company. This is forming a classic "Huang-style recruitment + Huang-style acquisition" model.
Beyond chips, NVIDIA is using talent acquisition to reshape its "second growth curve." In fiscal year 2025, NVIDIA's revenue reached $130.5 billion, more than doubling from the previous year, marking a miracle of growth in tech history. Concurrently, the company is actively rebuilding its secondary growth engine through strategic hiring.
On one front, it is systematically "poaching" talent to bolster critical capabilities in market strategy, policy, research, and organizational management. On another, it is acquiring startups to directly integrate core technical leaders and software骨干 into its ecosystem.
In terms of executive recruitment, NVIDIA brought in several new high-level managers over the past year, filling key roles in global markets and branding, human resources, quantum computing research, and cybersecurity policy.
In one of the latest personnel moves this year, NVIDIA aimed its recruitment shovel at Google. The chipmaker hired Google Cloud veteran Alison Wagonfeld as its first-ever Chief Marketing Officer (CMO).
Wagonfeld officially assumed the role in February, consolidating responsibilities previously scattered among several executives to take full charge of NVIDIA's marketing and communications.
(Note: NVIDIA had never before established a dedicated CMO position; related team leaders typically reported to a Vice President of Marketing rather than being managed uniformly by a single CMO.)
Following her appointment, all members of NVIDIA's marketing and communications team will report directly to Wagonfeld, who in turn reports to Jensen Huang.
Before joining NVIDIA, Wagonfeld served as Vice President of Marketing for Google and was the CMO for the core Google Cloud business unit. During her decade at Google, she was instrumental in building Google Cloud from a promising startup project in 2016 into a mature business with an annual revenue run rate of $60 billion.
This experience perfectly aligns with NVIDIA's current transition from "selling chips" to "selling systems and platforms." Reports indicate that Wagonfeld's mission at NVIDIA extends beyond external communications and sharing the burden of Jensen Huang's iconic leather jacket; it also involves helping the company establish a clearer, more sustainable market narrative among customers at all levels during the next phase of competition—shifting from training to inference.
On the human resources front, NVIDIA opted for a veteran from the traditional tech corporate system. HPE veteran Kristin Major joined NVIDIA in February of last year as Senior Vice President of Human Resources, becoming part of the executive leadership team directly overseen by Jensen Huang.
In human resources, Kristin Major is a genuine seasoned professional. Prior to NVIDIA, she worked at Hewlett Packard Enterprise (HPE) for over 13 years, holding various roles in HR and talent management, ultimately rising to Executive Vice President and Chief People Officer.
At HPE, she long managed multiple core business units within human resources, including GreenLake, Aruba Networking, and the HPE Transformation Office.
In the quantum computing domain, NVIDIA also cast its recruitment gaze toward Microsoft. Last November, NVIDIA poached Microsoft's key quantum computing figure, Krysta Svore, to serve as Vice President of Applied Research (Quantum Computing).
According to Krysta's LinkedIn, she will lead applied research and engineering work covering the entire quantum technology stack, focusing on advancing quantum error correction, system architecture, and AI-accelerated quantum workflows to accelerate the maturation of the quantum computing ecosystem.
Before NVIDIA, she worked at Microsoft for nearly 20 years, holding positions including Technical Fellow and Senior Vice President of Quantum Development, and was one of the key leaders behind Microsoft's quantum computing strategy. Her leadership involved integrating the first quantum computers into the Azure platform, driving advancements in quantum software and algorithms, creating an open-source quantum software technology stack, and designing scalable quantum architectures.
Furthermore, in a 2024 collaboration with Quantinuum and Atom Computing, she helped demonstrate logical qubits with error rates superior to physical qubits for the first time. This recruitment of Krysta Svore from Microsoft coincides perfectly with NVIDIA's accelerated phase of quantum布局. On one hand, NVIDIA is building a quantum research center; on the other, it is promoting the deployment of its open-source CUDA-Q platform in global quantum projects and collaborating with various quantum innovation institutions to bridge the path between classical and quantum computing.
Mark Weatherford joined NVIDIA in August 2025 as Head of Cybersecurity Policy and Strategic Partnerships.
Before NVIDIA, he held cybersecurity executive roles in both public and private sectors. He served as Chief Information Security Officer for the states of Colorado and California, and as Vice President and Chief Security Officer at the North American Electric Reliability Corporation (NERC), directly involved in developing and implementing cybersecurity standards for critical infrastructure in the power industry.
In the private sector, he handled cybersecurity strategy and policy at companies like Booking Holdings, Coalfire, and The Chertoff Group, and more recently held AI policy and standards-related positions at Gretel.
Of course, it's not a one-way street where Huang can only poach others' talent. In 2025, NVIDIA also saw the departure of several key executives.
Among them, former NVIDIA Senior Director of Robotics Research and head of the Seattle lab, Dieter Fox, left to join the Allen Institute for AI (Ai2) as Senior Research Director. His position has been taken over by NVIDIA Robotics Research Manager Yash Narang.
In the autonomous driving sector, former Vice President of Autonomous Driving Software and AI, Minwoo Park, moved to Hyundai Motor Group, assuming the role of Head of the Advanced Vehicle Platform (AVP) division and Company President, while also serving as CEO of the autonomous driving subsidiary 42dot.
Beyond poaching seasoned executives from giants like Google and Microsoft, another major personnel strategy for Jensen Huang in 2025 has been "acqui-hiring." This involves acquiring startups to directly absorb their core teams and key technologies, bringing talent, products, and roadmaps wholesale into NVIDIA.
Under this strategy, Huang shows a preference for teams that have proven their technology and engineering viability but haven't yet scaled up, with a significant portion hailing from Chinese entrepreneurs.
First, consider a typical "acqui-hire" case: to strengthen its布局 in AI Agents, enterprise generative AI, and efficient inference, NVIDIA completed the acquisition of Nexusflow last June. Joining as part of the deal were founder and CEO Jiantao Jiao, co-founder Banghua Zhu, CTO Jian Zhang, and other core members.
Post-acquisition, Tsinghua alumnus and 2011 undergraduate top award winner Jiantao Jiao became NVIDIA's Research Director, overseeing AI post-training, evaluation, agents, and related infrastructure. Fellow Tsinghua graduate Banghua Zhu became a Principal Research Scientist, and CTO Jian Zhang became Director of Applied Research, with Nexusflow's original technical core fully integrated into NVIDIA.
Prior to joining NVIDIA, Nexusflow completed a $10.6 million seed round in September 2023, reaching a post-money valuation of $53 million, with investors including Point72 Ventures and Fusion Fund. NVIDIA had previously strengthened collaboration through ecosystem connections like Together AI.
Acquired around the same time as Nexusflow was CentML, an acquisition aimed at enhancing the CUDA toolchain and model deployment efficiency, enabling developers to run models at lower cost and higher efficiency. Huang ultimately brought CentML into the fold for a total price exceeding $400 million. Like Nexusflow, this was a classic "acqui-hire."
Four co-founders, including post-95s Chinese CTO Shang Wang, and over 15 engineers were integrated into NVIDIA's system, joining its AI Software and Systems team directly. Specifically, former CentML CEO Gennady Pekhimenko became NVIDIA's Senior Director of AI Software, former COO Akbar Nurlybayev became Senior Manager of AI Software, Shang Wang took charge of AI Systems Software Management, and Chief Architect Anand Jayarajan also assumed an Engineering Manager role. Additionally, at least 18 other technical staff were transferred.
CentML, founded in 2022 and headquartered in Toronto, focused on AI model optimization software to improve GPU utilization and reduce training/inference costs. Its core technology is the Hidet compiler (a tensor compiler), which can automatically fuse operators, optimize scheduling, and leverage CUDA Graph to unlock GPU potential, reportedly boosting inference speeds by up to 8x, with support for PyTorch 2.0 integration.
Furthermore, to strengthen its布局 in cloud compute leasing, AI platforms, and vertical integration, NVIDIA also completed the acquisition of cloud service platform LeptonAI for several hundred million dollars last April. Post-acquisition, LeptonAI founder and former Alibaba VP Yangqing Jia joined NVIDIA as Vice President of System Software, responsible for overseeing the company's underlying system software and developer platform construction. Co-founder Junjie Bai also joined.
Lepton AI, founded in 2023 with an $11 million seed round, was dedicated to providing enterprises with an efficient, scalable AI application platform, allowing average developers to deploy AI models with just 2-3 lines of commands. The company launched a conversational search engine in December 2023 with less than 500 lines of code; its cloud GPU solution launched in June 2024 emphasized cost-effectiveness and reliability.
The acqui-hire of AI hardware startup Enfabrica is also noteworthy. Last September, NVIDIA, via an acqui-hire transaction valued at over $900 million, brought in Enfabrica founder and CEO Rochan Sankar and his core team, simultaneously obtaining related technology licenses.
Enfabrica, founded in 2019, developed technology capable of connecting over 100,000 GPUs into a unified computing system—a capability that fits perfectly into NVIDIA's关键 transition from "selling chips" to "selling complete machines and systems." Before acquisition, Enfabrica had completed multiple funding rounds, backed by industry capital from Arm, Samsung, Cisco, etc., representing a mature team with proven technology and engineering pathways.
In the open-source workload management arena, NVIDIA last year executed another classic "acqui-hire" by acquiring SchedMD, the primary developer of Slurm. Slurm is an open-source workload manager and job scheduler for High-Performance Computing (HPC) and AI, long used for managing task queuing, scheduling, and resource allocation in large-scale computing clusters, and is one of the core software components in current supercomputing and AI training infrastructure.
(Slurm already supports the latest NVIDIA hardware and is also one of the key infrastructures relied upon by generative AI, widely used by foundational model developers and AI builders to manage model training and inference tasks.)
Through this acquisition, SchedMD founder Danny Auble joined NVIDIA as Senior Director of System Software, continuing to lead system software work related to the Slurm direction.
The most recent "acquisition hiring" (more accurately, a technology licensing deal) occurred late last year, involving a transaction between NVIDIA and Groq. In December, NVIDIA hired Groq founder and TPU co-inventor Jonathan Ross and COO Sunny Madra, bringing along approximately 90% of the engineering team.
Prior to this, NVIDIA completed a $20 billion transaction with Groq, not only securing licensing rights to Groq's inference technology but also incorporating the core team. This deal is seen as a significant signal of the AI market's shift from the "training" to the "inference" phase.
Groq officially stated that despite the core team joining NVIDIA, the company would maintain independent operations, with former CFO Simon Edwards becoming CEO, and the cloud service platform GroqCloud continuing to offer services externally. This transaction also exemplifies NVIDIA's "acqui-hire" strategy: securing key technology while simultaneously吸纳 top talent, and providing a dignified exit for employees and shareholders—employees reportedly cashed out an average of around $5 million, and shareholders received payouts based on the $20 billion valuation.
As of now, the acqui-hire of Israeli star model company AI21 Labs is still underway, with an estimated acquisition price of $2-3 billion. Meanwhile, reports suggest NVIDIA is in talks to acquire domestic simulation and synthetic data leader Guang Lun Zhi Neng (Lightwheel AI) to complement the关键环节 in its Physical AI strategy—forming a more complete ecosystem闭环 from chips to platforms to data and services.
Overall, these frequent personnel changes clearly outline NVIDIA's strategic blueprint: By吸纳 top talent and acquiring high-growth startup teams, NVIDIA is rapidly transcending its identity as a mere GPU hardware supplier, building an integrated "hardware-software" full-stack platform at the system level. Its precise efforts in AI inference optimization, Agent deployment, and compute scheduling not only consolidate its dominance in existing markets but also, through提前布局 in quantum computing, data services, cybersecurity, and Physical AI, create interfaces ready for the next wave of AI红利.
This echoes Jensen Huang's "helmsman" declaration at CES 2026: NVIDIA is the "captain" of the AI industry, leading the global supply chain and partners.
Comments