Science And Technology Giants Gather Core, Chip Industry Chain Boom

Behind the development of chips by various tech giants are the inherent advantages of tech giants, as well as several key changes in the upstream of the chip design industry.

Core making is no longer the domain of chip companies

No matter Apple, Google, Microsoft, Tesla, Facebook, Amazon, or China's Baidu, Ali, Xiaomi, Vivo, have officially announced crossover layout chip design. Tencent and Bytedance have also been found to be recruiting chip-related positions.

Standardized chips don't necessarily deliver the most powerful performance when tech giants start chasing superior product performance.

Independent research and development can start from the hardware level, completely customized chip functions according to their own needs, realize their own software and hardware closely, so as to create the best products.

Apple outsourced chips to Samsung for the first three generations of its iPhones, but the continued underperformance prompted Apple to accelerate its own chip design team:

Google developed a TPU dedicated to the cloud because CPUs were struggling to meet the demands of its growing number of applications for accelerated AI neural network computing

With diverse requirements for future infrastructure, the traditional one-size-fits-all solutions are no longer appropriate.

Increasingly, these companies want to use custom chips, rather than the same ones as competitors, to meet the specific needs of their products and applications

It also gives them more control over software and hardware integration and helps them stand out from the competition.

On the other hand, its chips are cheaper than the market price. Products such as CPUs and special AI chips independently developed by tech giants often have a smaller footprint and lower power consumption than common chips in the market.

The savings in cost and square footage per chip may not be significant, but whether it is a data center server or a terminal product when the scale is large enough, it will bring considerable economic benefits.

Global computing is entering a golden age of architectural innovation

Due to the rise of heterogeneous computing, it is moving from general-purpose CPUs to parallel computing and distributed computing scenarios with Arm architecture processors, NPU and GPUs.

The increasingly perfect chip industry chain also lays a foundation for technology companies to make the core.

Over the past few years, the Arm architecture has been coming out of the generation and improving dramatically, bringing it abreast of the Intel-led x86 architecture.

By buying licenses from Arm, technology companies such as Huawei, Ikedance, Apple, Google, and Microsoft have the opportunity to design their own chips that are powerful enough.

Competing with chip giants for limited capacity

The rush to higher performance requires not only brilliant design but also advanced manufacturing techniques that can cram more transistors into the same area.

As only TSMC, Samsung and Intel are left in the world's advanced manufacturing process blockage players, Intel's external OEM business has just started and the capacity resources of advanced manufacturing process technology are quite limited.

Tech companies that choose to develop their own chips will have to compete for capacity with existing top chip makers such as Intel, Nvidia, and Qualcomm.

How to persuade OEM chipmakers to place their orders higher in the face of massive demand has become the primary challenge facing the tech giants.