Skip to main content Skip to main navigation Skip to search Skip to footer
Type to Search Subscribe View Tags

Transition in Semiconductor Foundries

Transition in Semiconductor Foundries
Abhishek Divekar - Global Engagement Manager | January 17, 2017
91 Views

The semiconductor industry lives by the simple creed of – smaller, faster, better, and as a result, is always in a state of constant flux. The disruptive ‘fabless –foundry’ business model, that evolved almost 4 decades ago, opened the doors for rapid growth of the industry in the last few decades. However, more recently, new dynamics have started impacting the industry.

The end of Moore’s law

For decades, the semiconductor foundries have been the torch bearers of the Moore’s law, implementing it religiously and pushing the boundaries of the wafer size to design increasingly smaller dies for circuit designs. The entire R&D of the foundries was focused on developing new technologies based on the Moore’s law. But the Moore’s law is now reaching it limits. The limits are not so much technical or physical as economical. To understand it, we need to understand how foundries function.

Fabless-Foundry Model

The foundries have been functioning based on the ‘fabless-foundry’ model. In the ‘fabless – foundry’ model, foundries are the pure play manufacturers while their customers, the fabless companies (& sometimes IDM’s) mostly focus on chip design. Foundries work on the principle of harnessing huge scales and high productivity to drive margins. They make large capital investments (upwards of $1 Billion) in building fabs to create newer technologies (smaller levels of wafer size). Then they engage with customers (usually the fabless companies) to drive the adoption of this new technology in their next chip design (known as ‘tape-outs’). As more customers shift to this new technology, foundries ramp up the production to harness economies of scale. Usually, a fab has to run at almost full capacity to generate positive margins. By the time this production is ramping up, they would have already invested in the R&D for the next technology. This cycle goes on.

Challenges

In the last few years, the fabless companies (foundry’s customers) have been facing challenges to frequently shift to the new technologies (especially sub 28nm). This is because the costs for designing chips for the smaller sized wafers are increasing disproportionately as compared to the resultant increase in the processing power from such chips (thus ending benefits from Moore’s law). Hence, fabless companies are showing reluctance to switch to the newer technologies. This is a cause of concern for the foundries as they are not able to ramp up the production of the new fabs quickly enough to reap returns on their investments before the next technology comes in. The challenge is further compounded due to the recent ramping up of the foundry capacities by some of the major IDM’s that are now competing with foundries.

Fabless-Foundry Model

Fig 1. The end of Moore’s law is near

Transition

To overcome this problem, foundries are transitioning to complete solution providers to support their customers to shift seamlessly to the new technologies. They are trying to develop capabilities beyond their traditional role of pure play manufacturing. These capabilities include product development capabilities like chip design, SoC design, prototyping, performance assessment, testing, layout design, packaging etc. However, developing these capabilities in house can be extremely difficult and time consuming. The pace at which the semiconductor industry functions does not allow such a leverage. Hence, foundries are looking towards avenues like acquisitions & partnerships to develop these capabilities quickly.  

To summarize,

The lines in ‘fabless – foundry’ model are blurring as we enter a new phase in the semiconductor industry

Consolidation will gather pace in the customer base of foundries, thus leading to fewer customers in the future. The dynamics based on which the industry thrived are about to change. Foundries must quickly adapt to these changes to stay competitive in the long run.

As an industry recognized leader in engineering services, HCL has a proven track record of partnering & supporting customers to overcome such challenges. 

References

Saha, S. (2016) Emerging Business Trends in the Microelectronics Industry. Open Journal of Business and Management4, 105-113. doi: 10.4236/ojbm.2016.41012

An Overview of the Semiconductor Industry http://marketrealist.com/2015/09/overview-semiconductor-industry/

TSMC annual report 2015 http://www.tsmc.com/download/ir/annualReports/2015_Business_Overview_E.pdf

http://www.wikinvest.com/industry/Foundry_Semiconductors

Image Source

http://www.wsj.com/articles/moores-law-runs-out-of-gas-1429282819


Contact Us
MAX CHARACTERS: 10,000

We will treat any information you submit with us as confidential. Please read our privacy statement for additional information.

We will treat any information you submit with us as confidential. Please read our privacy statement for additional information.

Sign in to Add this article to your Reading List
Register