Urgent.News

What's breaking now, across thousands of outlets.

AI

AI boom turns chip packaging into the next battleground for semiconductors

AI boom turns chip packaging into the next battleground for semiconductors

The global race to build AI infrastructure is driving investment in advanced chip packaging, a critical bottleneck in producing powerful processors. Companies like Nvidia, AMD and TSMC, the world's largest contract chipmaker, are expanding production. Packaging could eventually account for half the value of a chip product, with revenue from advanced packaging expected to surpass traditional chip manufacturing by 2026.

TSMC's CoWoS technology links separate chips on a shared base, enabling high-speed collaboration. Nvidia is financing Siliconware Precision Industries' expansion of CoWoS capacity, while AMD is investing over $10 billion in Taiwan's semiconductor industry. Alternative packaging solutions like panel-level technology, developed by Powertech, are being explored to reduce material waste and accommodate larger AI chips.

Chip designers are increasingly using advanced packaging to integrate multiple chips into a single system, overcoming the limitations of traditional packaging that mainly protected chips and provided electrical connections. The shift towards advanced packaging comes as it becomes harder and costlier to achieve performance gains by shrinking transistors.

Written by urgent.news from Investing.com's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at investing.com →

More in AI

Skymizer and MiTAC run 2 8 trillion parameter AI model on server

Technology vendors Skymizer and MiTAC Computing will showcase the 2 8 trillion parameter Kimi K3 artificial intelligence model running on a single air cooled server configured with two terabytes of…

  • Skymizer and MiTAC jointly unveil 2.8-trillion-parameter Kimi K3 model
  • Moonshot AI model runs on MiTAC server with eight HTX301 LPU cards
  • Collaboration aims to make open-weight models practical for on-premises deployment

More from Friday 6 November →