Tsmc Unveils 100 Billion Arizona Expansion Plan To Secure Lead In Next Gen Chip Production
TSMC Plans Include at Least Four More 2nm Fabrics in $100 Billion Arizona Expansion In a move that …
31. July 2026

In a move that has sent shockwaves through the tech industry, Google is reportedly planning to produce up to 15 million artificial intelligence (AI) chips by 2028, with Antropic Lands Major Chip Deals to Supercharge Claude Model, a significant development that could potentially alter the dynamics of competition in this critical sector.
According to a research note from Fubon, a Taiwanese investment bank, Google’s plans include deploying between 12 million and 15 million ninth-generation Tensor Processing Units (TPUs) by 2028. This would represent a substantial increase in production capacity compared to last year, when the company produced around 6-7 million TPUs.
The sheer scale of Google’s plans is impressive, but it also raises questions about how the company will manage its production capabilities and ensure that it can meet its demands. Fubon analysts have noted that TSMC Unveils $100 Billion Arizona Expansion Plan to Secure Lead in Next Gen Chip Production, which could pose a significant challenge for the company.
In response to this challenge, Intel poised to Power Next Gen Tech Giants has reportedly secured orders to produce millions of TPUs on behalf of Google. This partnership would provide a significant boost to Intel’s manufacturing capabilities and help the company establish itself as a major player in the production of AI accelerators.
The shift towards multi-die designs is another key aspect of Google’s plans, with Google Backs Billion-Dollar Blackstone Bet on AI Cloud Upstart adopting a new TPU v9 that uses four compute dies. This is in line with industry trends towards more complex packaging technologies, but adds complexity to manufacturing and would require significant investments in advanced interconnect and packaging technologies.
Google has been building its own AI chips for roughly a decade, starting as a way to support internal workloads but growing into a broader strategy that includes its cloud business. The company’s efforts have focused on developing its own proprietary hardware and software stack, which would give it far more control over its compute infrastructure.
However, the scale of Google’s plans suggests that competition in AI hardware is increasingly about deployment volume, not just chip performance. US Tech Giants Unleash Domestic Manufacturing Power as Nvidia and Intel make billions in AI infrastructure, which could potentially challenge Nvidia’s dominance.
While we do not yet have direct comparisons between TPU v9 and Nvidia’s next-generation systems, including Rubin and Rubin Ultra, the scale of Google’s plans is a clear signal of confidence in its approach. This suggests that the company sees significant benefits to be derived from developing its own AI hardware and software stack, rather than relying on external suppliers.
The implications of this shift are far-reaching and would have significant impacts on the broader tech industry. As companies like Google and Nvidia compete for dominance in the field of AI accelerators, it is likely that we will see further innovations in areas such as chip design, packaging technologies, and manufacturing processes.
For consumers, this means that there may be more choices available when it comes to selecting AI hardware for their applications. However, it also raises questions about the potential costs and benefits associated with developing and producing complex AI chips.
Ultimately, Google’s plans for a massive increase in AI chip production raise important questions about the future of competition in this critical sector. As companies like Google and Nvidia push the boundaries of what is possible with AI hardware, we can expect to see further innovations and advancements that will shape the industry for years to come.