Nvidia lifts buyback ceiling to $235 billion with new $150B tranche
Macro & Markets ·
Nvidia's board has cleared an additional $150 billion for share repurchases, pushing the company's cumulative buyback authorization to $235 billion as demand for AI computing hardware continues to drive its financial results.
The new authorization marks a substantial increase in the capital Nvidia has set aside to repurchase its own stock, a move companies typically use to return excess cash to shareholders and support share value. Because buybacks reduce the number of outstanding shares, a program of this scale signals confidence from Nvidia's leadership that the company will keep generating cash well beyond its operating needs, even as it continues to invest heavily in chip production and research.
The decision comes against the backdrop of extraordinary revenue growth tied to the artificial intelligence buildout. Nvidia's most recent fiscal first quarter brought in $81.62 billion, up 85 percent from the prior year, with the bulk of that total coming from its Data Center segment, which sells the training and inference chips that cloud providers, enterprises, and governments rely on for AI workloads. That trajectory has made Nvidia central to the infrastructure underpinning generative AI systems built on its GPU architecture and CUDA software platform.
The buyback expansion is not happening in isolation. Related developments in Nvidia's ecosystem point to sustained demand for its hardware across new markets: Bitdeer has shifted into AI infrastructure with plans to build what would be Norway's largest data center to support next-generation Nvidia chips, while a firm called Ornn has introduced forward-pricing contracts for Nvidia's H100, H200, and B200 GPUs, aiming to bring market-based transparency to future AI compute supply and demand. Nvidia has also introduced Ising, described as its first open AI models for quantum computing, intended to support scalable systems through AI-assisted calibration and error correction.
Not every signal in the surrounding coverage is unambiguously favorable. Nvidia's chief executive has said China already possesses the compute capacity to train AI systems at a level comparable to Claude Mythos, a comment that has raised cybersecurity concerns. Separately, Google is reported to be in talks with Marvell to develop next-generation AI chips, including a new memory processing unit and TPU design, in an effort to challenge Nvidia's position in the GPU market. Details on the timing and structure of the new $150 billion repurchase program, including how quickly Nvidia intends to execute it, have not been disclosed. The buyback authorization itself is tracked as part of the broader Buybacks trend across corporate capital-return activity, alongside ongoing questions about how competitive pressure from custom silicon and export restrictions on chips to China may shape Nvidia's growth going forward.