Google Unveils New AI Chips in Major Challenge to Nvidia

Google is stepping up its efforts to challenge Nvidias dominance in the intelligence chip market by introducing new specialised processors and strengthening Google’s hardware partnerships. The Alphabet‑owned company is moving beyond its two‑year product launch cycle to speed up Google’s chip development and meet the rapidly growing demand, for AI computing.

The company announced versions of its custom‑built Tensor Processing Units or TPUs. These new TPUs are designed to meet computing needs instead of using one single design, for all tasks. This change shows that Google wants to become a name in the very competitive AI hardware industry.

New TPU 8tx and TPU 8i Max chips

Google has introduced two new chip variants: the TPU 8tx and TPU 8i Max. According to Amin Vahdat, Google's head of chip development, specialisation will play a crucial role in shaping the future of AI computing.

The TPU 8tx focuses on security and confidential computing. The TPU 8tx is designed for government agencies and businesses operating in regulated industries where protecting sensitive information is a priority. The TPU 8tx aims to keep data isolated and secure while handling demanding AI workloads.

Meanwhile, the TPU 8i Max targets applications that require significant computing power. Google has combined two TPU chips into a single unit, offering double the memory and processing capabilities. This makes the chip suitable for handling large datasets and complex AI operations, including advanced mathematical calculations, software development and drug discovery.

Google strengthens ties with AMD

Google has also partnered with Advanced Micro Devices (AMD) to support its security-focused TPU 8tx platform. The arrangement combines Google's specialised AI technology with an AMD processor, helping ensure that sensitive data and AI models remain encrypted at the hardware level during processing.

The partnership is particularly interesting given AMD CEO Lisa Su's family connection with Nvidia CEO Jensen Huang, who are cousins. Google's decision to work with AMD adds another dimension to the competition against Nvidia.

Lower costs attract major technology companies

Pricing is emerging as one of Google's biggest advantages. Market analysts estimate that its TPUs cost around 30 per cent less than Nvidia's popular graphics processing units (GPUs). They also offer competitive processing speeds while consuming less power.

Several major technology companies, including Apple, Meta, OpenAI and Oracle, are reportedly interested in Google's hardware. Anthropic, the company behind Claude, reportedly has an option to purchase up to $150 billion worth of chips between 2028 and 2031. Financial firm Jump Trading is also adopting TPUs to train AI models.

Revenue projections and supply challenges

According to Bloomberg Intelligence, Google's TPU sales could grow from approximately $12 billion this year to more than $135 billion by 2028.

However, competition from startups such as Tenspresso, founded by former Google engineers, remains a challenge. Analysts also anticipate a possible 5 to 7 per cent slowdown in global data centre infrastructure spending by 2027.

Google is additionally facing shortages of memory components and manufacturing capacity. With every TPU currently contracted to a buyer, the company is exploring partnerships with Marvell Technology and MediaTek to diversify its supply chain and avoid production bottlenecks.

ABOUT THE AUTHOR

Kahekashan

[email protected]

Kahekashan is a passionate technophile with a keen eye for cutting-edge gadgets, emerging technologies, and everything in the digital realm. Raised in a Defence family with strong values and a background in literature, she has consistently pursued excellence in every endeavour. Her last full-time assignment involved content writing with the Indian School of Business.

Read More