open_in_newOriginal article
Google is reportedly switching from Samsung to TSMC for the Tensor G5. Why could this be a big deal?

Google is reportedly switching from Samsung to TSMC for the Tensor G5. Why could this be a big deal?

  • 20.03.2025 19:07
  • msn.com
  • Keywords: Switching chip manufacturers, Potential performance improvements

Google is switching Tensor G5 chip production from Samsung to TSMC, which uses a more advanced 3nm process. This move could enhance performance and quality, with Google designing more custom components like the ISP for better camera processing.

Alphabet News

Estimated market influence

Google

Positivesentiment_satisfied
Analyst rating: N/A

Google is switching to TSMC for Tensor G5 chips, which could lead to better performance and quality.

Samsung

Negativesentiment_dissatisfied
Analyst rating: N/A

Samsung's issues with chip manufacturing led Google to switch to TSMC.

TSMC

Positivesentiment_satisfied
Analyst rating: N/A

TSMC's advanced manufacturing processes may improve Tensor G5 performance and reliability.

Context

Analysis of Google Switching from Samsung to TSMC for Tensor G5

Key Facts and Data Points

  • Current Fabrication Partner: Samsung Foundry (for Tensor G1-G4)
  • New Fabrication Partner: Taiwan Semiconductor Manufacturing Company (TSMC) (starting with Tensor G5)
  • Process Node Upgrade: Transition from 4nm to TSMC's 3nm process node
  • Expected Benefits:
    • Improved performance and quality due to TSMC's advanced manufacturing capabilities
    • Better yield rates compared to Samsung Foundry's reported issues with smaller nodes

Market Implications

Performance and Quality Improvements

  • Process Node Upgrade: The move to TSMC's 3nm process node is expected to deliver significant performance and efficiency gains.
  • Image Signal Processor (ISP): Tensor G5 will feature a Google-designed ISP, potentially leading to better camera output despite similar hardware.

Competitive Dynamics

  • TSMC's Leadership: TSMC is the industry leader in semiconductor manufacturing, producing chips for major players like Apple (e.g., Apple A18 Pro).
  • Samsung Foundry's Challenges: Samsung has faced issues with 4nm and smaller node yields, which could limit its competitiveness in high-end chip production.

Strategic Considerations

  • Design Flexibility: Google is replacing Samsung-made controllers (e.g., USB, display) with licensed parts from companies like Synaptics, offering more flexibility in hardware optimization.
  • Custom Components: Tensor G5 will include next-generation Tensor Processing Units (TPUs) and Digital Signal Processors (DSPs), enhancing AI capabilities.

Long-Term Effects

  • Supply Chain Diversification: The switch to TSMC reduces reliance on Samsung Foundry, potentially improving reliability and cost efficiency.
  • Potential Cost Savings: Better yield rates from TSMC could lead to lower production costs for Google.

Unchanged Aspects

  • GPU and CPU Architecture: Tensor G5 will still use Imagination Technologies DXT GPU and Arm Cortex CPUs, similar to previous generations.
  • Core Features: Audio processor, memory controller, and other components remain unchanged, indicating incremental rather than revolutionary updates.

Conclusion

The switch from Samsung Foundry to TSMC for the Tensor G5 represents a strategic move by Google to leverage TSMC's advanced manufacturing capabilities, potentially improving hardware quality and performance. While not a complete overhaul, this change could mark a turning point for Pixel hardware optimization and competitiveness in the smartphone market.