Google Samsung partnership targets next-gen AI chip as tech giant moves to diversify suppliers

4 Sources

Share

Google is negotiating with Samsung Electronics to manufacture a critical component of its upcoming Tensor Processing Unit, codenamed Icefish. The deal would use Samsung's advanced 2-nanometer technology for the memory-connecting component, while TSMC handles the main computing portion. Mass production could begin in 2028 as Google works to reduce its dependence on a single chipmaker.

Google AI Chip Strategy Takes Shape with Samsung Talks

Google is in discussions with Samsung Electronics to manufacture a component of its next-generation AI processor, marking a significant shift in the tech giant's chip manufacturing strategy

1

. The Icefish processor, Google's tenth-generation Tensor Processing Unit, is being developed in partnership with chip design firm MediaTek and could enter mass production as early as 2028

2

. Under the proposed arrangement, TSMC would manufacture the main computing portion of the chip, while Samsung would produce a memory-connecting component using its advanced Samsung 2-nanometer technology

3

. The 2nm process packs more power into smaller chips, potentially improving speed, power use and AI capabilities.

Source: DT

Source: DT

Samsung's Push to Expand Contract Manufacturing Business

Landing the Google Samsung partnership would represent a major win for Samsung's efforts to grow its contract chip-manufacturing business in the semiconductor industry

3

. Samsung has been actively pursuing customers for its advanced manufacturing processes, announcing in April that it expects to win more customers and was considering a second Texas plant to ramp up production

1

. The company already secured a $16.5 billion deal with Tesla in July 2025 to make AI chips using the same 2-nanometer process

3

. This momentum signals Samsung's determination to compete more aggressively in the foundry market against industry leader TSMC.

Efforts to Reduce Reliance on TSMC and Diversify Chip Manufacturing

The discussions reflect Google's broader strategy to reduce reliance on TSMC, which is struggling to keep up with surging AI demand and could prove to be a bottleneck for the industry

1

. This move comes amid significant strain on global semiconductor production capacity, prompting major technology players to diversify chip manufacturing sources

4

. Google is also in talks with Intel to manufacture more than three million TPUs in 2028, further illustrating the company's determination to secure its long-term supply chain

3

4

. By spreading production across multiple partners, Google aims to mitigate risks associated with depending on a single manufacturer in the global chip war.

Source: ET

Source: ET

Tensor Processing Units Challenge Nvidia's Market Position

Google's in-house AI chip capabilities have emerged as a viable alternative to Nvidia's dominant graphics processors, with rising sales of Tensor Processing Units becoming a growth driver for its cloud unit

1

. The company unveiled two new custom chips in April, designed for training AI models and inference

3

. These specialized chips are designed to train and run artificial intelligence models, becoming increasingly important for Google's cloud business and AI services

2

. Industry experts believe the Google AI chip development reflects a much larger shift taking place in the global AI and semiconductor markets, as companies seek to break free from dependence on Nvidia's ecosystem

2

. The Icefish processor is still in the design phase, with MediaTek collaborating on the development of this next-generation AI processor

4

.

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved