Intel Finds a New Door Into the AI Chip Boom Through Advanced Packaging

There is an indication that Intel could discover an alternative way to grow quickly in its foundry operations. In addition to focusing on its efforts to manufacture silicon wafers, the firm wishes to make packaging of advanced chips one of its main sources of income.

In accordance with South Korean press reports, there are a number of firms including Broadcom, MediaTek and Meta Platforms which have an interest in packaging offered by Intel. In addition, there could be an interest in the technology by Google via collaboration with Broadcom in TPU processors.

It is essential to note the moment when this is happening. There is a huge demand for CoWoS packaging offered by TSMC. Customers have even booked production slots a few years in advance.

EMIB Targets the Next Generation of AI Hardware

The EMIB solution, offered by Intel, serves as an alternative for the well-known CoWoS platform by TSMC. With EMIB, chipmakers are able to integrate various elements into a single big package, with the overall size of the chip reaching 120 by 120 mm.

Furthermore, there is EMIB T, which integrates vertical connections, thus improving the performance of power distribution and making signal paths shorter. This can prove useful for developing increasingly sophisticated AI accelerators.

Advanced packaging solutions are expected to make a notable contribution to foundry revenue starting the second half of next year.

Importance of it for the Market

In our opinion, there is no need for Intel to win in all fronts to be in a good position. It needs only to become a second source in the market which suffers from a shortage of packaging capacities.

For you, it means that you may get some more competition, more AI hardware and less bottlenecks. It can also mean some greater flexibility in design options for chip makers.

If you find our analysis interesting, please do us a favor by liking and sharing it on your favorite social network. You can also check out other articles related to semiconductors, AI hardware and future technologies of computing.