News Article
What’s driving the R&D Outsourcing Boom
Semiconductor design remains a growth area for R&D outsourcing as the benefits to businesses become more obvious. R&D, it appears, is no longer the sacred cow that it once was, something companies would never share with other companies as reported by Wipro
Why is there a sudden increase in Research and Development (R&D) outsourcing?
The semi-conductor market is extremely feature driven, and companies today realise that closed innovation may not be the right way forward. Many companies are looking outwards to generate new ideas - ideas that will focus on core differentiation in products. This is as true for the emerging MEMS market.
What are the benefits of R&D outsourcing?
Frees internal resources for "value-add" tasks The most immediate benefit of outsourcing is that it allows companies to focus critical attention on their core activities, leaving the non core activities to partners who they are confident will both understand their business and be with them for the long haul.
Another reason that companies look at outsourcing some of their R&D is to fill in gaps in their expertise. With device and end system complexities increasing by the day, it is difficult for a single organisation to be able to build all the necessary skills in-house. Far better to leverage organisations that have strong skills and Intellectual Property (IP) in a particular domain, for that part of the solution. Another area open to outsourced R&D is integration of the entire solution where, again, there are firms which specialise by developing a good understanding of the different end-user markets, the various components and Intellectual Property available from vendors and in the public domain.
Reduce operating costs for productivity gains
Another benefit that companies see in outsourcing is reducing the total cost of that development. Owing to the labour cost arbitrage (skilled-labour wage discrepancies), certain geographies are able to offer lower operating costs for setting up development centres. But with the global shortage of talent, no geography is insulated from cost escalations, and this low cost advantage might not be sustainable by them in the long term. Sensible companies have therefore invested heavily in increasing their overall productivity and quality levels significantly and are able to offer better productivity per dollar.
There is compelling evidence to suggest that more companies are buying into the idea of handing over the entire R&D process to partners who may possess better processes and therefore be more efficient than themselves.
With Asia Pacific increasingly becoming a very important market, a lot more companies are realising the need to have a presence closer to their end-user customers and this has also significantly contributed towards the new wave of R&D outsourcing to destinations like India and China. The shortage of sufficiently skilled personnel in the geography where the company currently operates means that companies have a real need to access the global talent pool wherever it may be.
Companies have a choice to setup a centre in a new geography or to outsource work to a vendor who is able to tap that talent pool more easily. More often than not, the costs and risks of operating an independent centre in a new geography is fairly high, especially for a medium sized organisation. This drives them towards an external vendor who can provide R&D services from start to finish. The strongest case for companies to outsource is to cut the "time-to-market". As product cycles shrink, companies are under pressure to launch products faster and faster. As a result of this, companies, which aim to reduce their product cycle time by as much as 50%, realise that there is a need to work with partners and deliver results within the shortest possible timeframe.
One example of reducing cycle time is to use the time zone differences to advantage by running a 24x7 operation for certain engineering activities. Though not all activities are amenable to such a model, there are quite a few in the engineering cycle such as testing, regression cycles, and some innovative implementations of reviews which allow an organisation to effectively use the full 24 hours. Outsourced product development can save a company US$2-5 million in costs, but outsourced R&D - even if a product hits the market three months earlier - could lead to US$80-100 million in additional revenues, since time-tomarket is crucial to product companies,
The Microelectronics Industry
In trying to understand the potential of outsourcing in the MEMS market it is important to understand the process of outsourcing in the semiconductor industry.
There are three primary activities in the value chain:
- Design
- Fabrication (more commonly referred to as "front end")
- Test and Assembly (more commonly referred as "back end").
Just like other industries, the semiconductor industry has also undergone successive phases of outsourcing, starting first with assembly, then fabrication and finally design with each phase marked by its own unique economic characteristics.
1) Assembly: The first activity in the value chain that was moved offshore was assembly since it was functionally separate from the other phases. Assembly required test and assembly equipment which were expensive, but the overall costs of plant and equipment were much lower than what was needed for a fully fledged fab (fabrication) which was to the tune of US$ 2-3 billion. Assembly also began with a relatively high use of less-skilled labour which was slowly replaced with automation.
2) Foundries: The next phase was the establishment of ‘pure play foundries' which were mainly companies engaged in offshore fabrication of chips for their customers. These foundries simply manufactured chips to the designs of other companies (commonly referred to as fabless or ‘design-only' chip makers) and sold no chips of their own design.
At the same time, there were IDMs (Integrated device manufacturers) who used their own fabs fully and who used the foundries more for buffer capacity thereby hedging the enormous risk of building new factories. As the pressure on companies to hit the market with products earlier than their competitors mounted, many companies realised the value of designing to a 24-hour cycle, as a result of which design outsourcing became a fast emerging trend. In addition, companies also licensed standardised portions or reusable modules known as IP blocks or "cores" to save time. While issues like IP protection and confidentiality still persist, design outsourcing as a phenomenon has been well received and will be driven by the continual need to reduce rapidly rising costs in order to keep expanding demand, especially in consumer markets.
3) Design: The ‘design' phase for semiconductor companies today also means a lot of supporting software development. Depending on the silicon that is being designed, it could range from supporting diagnostics, device drivers, different Operating Systems availability and porting, multimedia codec implementation and more.
The demand for ready-to-run software on the silicon is very high especially with the product design cycles becoming very short and the product design companies not having a large software team to handle this task. The extreme competition in certain market segments among the semiconductor players forces them to give away this supporting software almost for free, despite the fact that it is a large volume of activity for them.
So, another key area where companies are looking for help is in reducing the cost of developing all the supporting software and they are looking at options such as outsourcing to help them achieve this goal. The additional benefit is that these design houses will then have a good knowledge base on the particular silicon, and there is a high probability that their design teams would pick that silicon over other comparable parts when they are building an end system for another customer.
The semi-conductor market is extremely feature driven, and companies today realise that closed innovation may not be the right way forward. Many companies are looking outwards to generate new ideas - ideas that will focus on core differentiation in products. This is as true for the emerging MEMS market.
What are the benefits of R&D outsourcing?
Frees internal resources for "value-add" tasks The most immediate benefit of outsourcing is that it allows companies to focus critical attention on their core activities, leaving the non core activities to partners who they are confident will both understand their business and be with them for the long haul.
Another reason that companies look at outsourcing some of their R&D is to fill in gaps in their expertise. With device and end system complexities increasing by the day, it is difficult for a single organisation to be able to build all the necessary skills in-house. Far better to leverage organisations that have strong skills and Intellectual Property (IP) in a particular domain, for that part of the solution. Another area open to outsourced R&D is integration of the entire solution where, again, there are firms which specialise by developing a good understanding of the different end-user markets, the various components and Intellectual Property available from vendors and in the public domain.
Reduce operating costs for productivity gains
Another benefit that companies see in outsourcing is reducing the total cost of that development. Owing to the labour cost arbitrage (skilled-labour wage discrepancies), certain geographies are able to offer lower operating costs for setting up development centres. But with the global shortage of talent, no geography is insulated from cost escalations, and this low cost advantage might not be sustainable by them in the long term. Sensible companies have therefore invested heavily in increasing their overall productivity and quality levels significantly and are able to offer better productivity per dollar.
There is compelling evidence to suggest that more companies are buying into the idea of handing over the entire R&D process to partners who may possess better processes and therefore be more efficient than themselves.
With Asia Pacific increasingly becoming a very important market, a lot more companies are realising the need to have a presence closer to their end-user customers and this has also significantly contributed towards the new wave of R&D outsourcing to destinations like India and China. The shortage of sufficiently skilled personnel in the geography where the company currently operates means that companies have a real need to access the global talent pool wherever it may be.
Companies have a choice to setup a centre in a new geography or to outsource work to a vendor who is able to tap that talent pool more easily. More often than not, the costs and risks of operating an independent centre in a new geography is fairly high, especially for a medium sized organisation. This drives them towards an external vendor who can provide R&D services from start to finish. The strongest case for companies to outsource is to cut the "time-to-market". As product cycles shrink, companies are under pressure to launch products faster and faster. As a result of this, companies, which aim to reduce their product cycle time by as much as 50%, realise that there is a need to work with partners and deliver results within the shortest possible timeframe.
One example of reducing cycle time is to use the time zone differences to advantage by running a 24x7 operation for certain engineering activities. Though not all activities are amenable to such a model, there are quite a few in the engineering cycle such as testing, regression cycles, and some innovative implementations of reviews which allow an organisation to effectively use the full 24 hours. Outsourced product development can save a company US$2-5 million in costs, but outsourced R&D - even if a product hits the market three months earlier - could lead to US$80-100 million in additional revenues, since time-tomarket is crucial to product companies,
The Microelectronics Industry
In trying to understand the potential of outsourcing in the MEMS market it is important to understand the process of outsourcing in the semiconductor industry.
There are three primary activities in the value chain:
- Design
- Fabrication (more commonly referred to as "front end")
- Test and Assembly (more commonly referred as "back end").
Just like other industries, the semiconductor industry has also undergone successive phases of outsourcing, starting first with assembly, then fabrication and finally design with each phase marked by its own unique economic characteristics.
1) Assembly: The first activity in the value chain that was moved offshore was assembly since it was functionally separate from the other phases. Assembly required test and assembly equipment which were expensive, but the overall costs of plant and equipment were much lower than what was needed for a fully fledged fab (fabrication) which was to the tune of US$ 2-3 billion. Assembly also began with a relatively high use of less-skilled labour which was slowly replaced with automation.
2) Foundries: The next phase was the establishment of ‘pure play foundries' which were mainly companies engaged in offshore fabrication of chips for their customers. These foundries simply manufactured chips to the designs of other companies (commonly referred to as fabless or ‘design-only' chip makers) and sold no chips of their own design.
At the same time, there were IDMs (Integrated device manufacturers) who used their own fabs fully and who used the foundries more for buffer capacity thereby hedging the enormous risk of building new factories. As the pressure on companies to hit the market with products earlier than their competitors mounted, many companies realised the value of designing to a 24-hour cycle, as a result of which design outsourcing became a fast emerging trend. In addition, companies also licensed standardised portions or reusable modules known as IP blocks or "cores" to save time. While issues like IP protection and confidentiality still persist, design outsourcing as a phenomenon has been well received and will be driven by the continual need to reduce rapidly rising costs in order to keep expanding demand, especially in consumer markets.
3) Design: The ‘design' phase for semiconductor companies today also means a lot of supporting software development. Depending on the silicon that is being designed, it could range from supporting diagnostics, device drivers, different Operating Systems availability and porting, multimedia codec implementation and more.
The demand for ready-to-run software on the silicon is very high especially with the product design cycles becoming very short and the product design companies not having a large software team to handle this task. The extreme competition in certain market segments among the semiconductor players forces them to give away this supporting software almost for free, despite the fact that it is a large volume of activity for them.
So, another key area where companies are looking for help is in reducing the cost of developing all the supporting software and they are looking at options such as outsourcing to help them achieve this goal. The additional benefit is that these design houses will then have a good knowledge base on the particular silicon, and there is a high probability that their design teams would pick that silicon over other comparable parts when they are building an end system for another customer.
