Could Google Benefit From Nvidia Leadership Shakeup? OpenAI, Anthropic Mega Orders Extend TPU Visibility to 2029
Nvidia’s leadership transition provides Google a strategic window to expand its custom TPU business. Analyst Ming-Chi Kuo reports that supply visibility extends through 2029, driven by substantial orders from OpenAI and Anthropic totaling approximately 26 million TPUs for 2028 and 2029. While Nvidia maintains product advantages, management uncertainty prompts major clients to diversify supply chains. Google plans to supply older v8 series variants to external partners while focusing internally on v9 chips. This demand indicates that the custom chip investment cycle will likely outlast market concerns regarding a 2028 capital expenditure peak, strengthening Google's commercial alternative.

TradingKey - Nvidia's (NVDA) management changes are creating a new window of opportunity for Google (GOOGL) to expand its AI chip business. Prominent supply chain analyst Ming-Chi Kuo noted that recent increases in custom chip orders from OpenAI and Anthropic have extended shipment visibility for Google's TPUs through 2029, ahead of most Nvidia GPUs and other AI ASIC suppliers.
According to Ming-Chi Kuo's supply chain survey, OpenAI and Anthropic are expected to require a combined total of approximately 26 million TPUs between 2028 and 2029. Of this total, OpenAI's demand is around 12 million units, primarily through direct purchases, while Anthropic's demand is about 14 million units, combining chip procurement and compute leasing.
Nvidia's product advantages will not disappear immediately as a result, but Jensen Huang's resignation and the fact that a permanent successor has not yet been determined could prompt customers to place greater emphasis on supply diversification when formulating compute plans for 2028 and beyond. For Google, this represents a key opportunity to transition its TPUs from internal infrastructure into an external commercial product.
OpenAI and Anthropic Push TPU Orders to 2029
The TPU is a proprietary AI accelerator developed in-house by Google, initially used primarily for internal workloads across Search, advertising, Gemini, and Google Cloud. Unlike GPUs designed for diverse computing tasks, the TPU is engineered specifically for machine learning training and inference, optimizing performance, power consumption, and total cost of ownership for specific AI models.
Ming-Chi Kuo expects Google to begin shifting its computing power expansion focus to the v9 TPU in 2028. Meanwhile, the earlier v8 series and its design variants will primarily be supplied to external customers such as OpenAI and Anthropic, thereby extending the commercial lifecycle of older-generation products.
The approximately 12 million chips that OpenAI plans to procure primarily correspond to the 8t series co-designed with MediaTek. The roughly 14 million chips required by Anthropic are mainly the 8i series co-designed with Broadcom. Demand from both customers is expected to continue increasing in 2028 and 2029.
However, supply chain capacity may constrain actual delivery volumes, leading Ming-Chi Kuo to estimate Google TPU shipments from 2027 to 2029 at 8 million to 9 million units, 14 million to 15 million units, and 16 million to 18 million units, respectively.
The most notable takeaway from this forecast is that TPU shipments are still expected to grow in 2029. The market had previously been concerned that Big Tech AI capital expenditures might peak around 2028, but orders from OpenAI and Anthropic suggest that the investment cycle in custom chips by frontier model companies may be longer than expected.
Why Nvidia Leadership Change Offers Google an Expansion Window?
Following Jensen Huang's resignation as Nvidia's chief executive officer and board director, Chief Financial Officer Colette Kress was appointed interim CEO, while the board searches internally and externally for a permanent successor. Although the company stated that its existing product roadmap and customer commitments would remain uninterrupted, the market continues to assess the long-term impact of the leadership change on R&D, supply chains, and customer relationships.
Ming-Chi Kuo believes that uncertainty surrounding the succession plan could help Google discuss long-term procurement plans beyond 2028 with customers. For enterprises preparing to invest tens of billions of dollars in building data centers, a supplier's product roadmap, management stability, and delivery capabilities over the coming years are just as crucial as single-chip performance.
Nvidia's historical competitive edge stemmed not only from GPUs and CUDA, but also from Jensen Huang's deep involvement in TSMC's advanced process nodes, CoWoS packaging, HBM supply, and capacity allocation for major customers. During periods of tight capacity, these relationships helped Nvidia secure critical resources ahead of time.
As the company enters a leadership transition, OpenAI and Anthropic may seek to reduce their future computing plans' reliance on a single company and a single key executive. With its own chip architecture, cloud platform, data centers, and AI models, Google can provide customers with an integrated alternative spanning from hardware to cloud services.
This content was translated using AI and reviewed for clarity. It is for informational purposes only.
Recommended Articles













Comments (0)
Click the $ button, enter the symbol, and select to link a stock, ETF, or other ticker.