Semiconductor Socket Wars – III

Published November 5, 2015

Other War Types

Apart from the Invader-Defender type of Socket war that was discussed intensively in the prequel , there are two other war-types. Namely:

  1. War for the new world: when the application or design is totally new and all sockets are up for bid.
  2. Alien Attack: When a chip is not attacked by a direct competitor but instead the socket itself is swallowed by a chip in another socket.

These wars are not fought every day, but when they happen they change the business environment for good.

War For The New World

In the war for the new world there is no defender – all the players are invaders charging in for the gold mine! A “new world” scenario evolves, usually due to one of the following two conditions:

  1. Customers building a new application/product
  2. Changing the design of a traditional application/product

The war for the new world is not just different because of the absence of a defender – it could be even more complex since the sockets itself may not be pre-defined. Usually when the customers are willing to change the traditional design they are keen to explore multiple alternatives and eventually choose the one which is aligned the most with their overall business objective.

It would be useful to understand the full picture here, using the following matrix:

 

 Product/ Application Type – NewProduct/ Application Type – Traditional
Design type –New

The most complex of the war-types but the results could be most exciting. Usually in such a case customers are willing to innovate in every aspect of the project and are keen to listen to the chip-makers on all the alternates before finalizing a design.

A great place to be in for the most proprietary and innovative chips that are struggling hard to enter traditional sockets.

To win the war you need to convince the customer on your value proposition, train feverishly and work with the engineering team closely to realize the final design.

If you are a start-up or a relatively niche player – top management involvement in such businesses makes sense.

Reinventing the wheel – maybe to roll it faster.

When the customer is redesigning an existing product/application their intent could be manifold –

  1. Make the design more cost effective.
  2. Enhance the product to meet the new age requirement
  3. Other barriers w.r.t. adopting the traditional design (e.g. IP protection etc.)

You cannot fight this war successfully without knowing the intent. Figuring out the “customer’s intent” should be the focus of the initial discovery meetings. Based on the understanding, the plan of action then needs to be drafted.

Design type – traditional

The old boy in a new shirt.

All sockets are pre-defined and the traditional socket war ensues with the single exception of there being no defender.

Usually when the customer is adopting a traditional design for a new product or application they are doing so to leverage their existing supplier relations, mitigate risks and to save on investments (time and money).

In such a war you are likely to meet your old enemies and hence the strategies would get predictable. Fighting it out the traditional Invader-Defender way makes more sense.

In this case the customer is likely to be a “me-to” late entrant to an existing market space.

Usually the intent of the customer here is to ride on the existing demand of a product. However being a late entrant he is aware that he would face tremendous margin pressure – prices are already rationalized and competitors would have better costs due to economy of scale.

In such a case it is likely that the odds would be favorable for the lowest cost supplier. In other words such a war would be a price-war (with your usual competitors) which you may choose to fight or walk-away from.

 

ALIEN ATTACK

An ‘alien attack’ happens when a socket disappears due to a chip housed in a different socket on the board. Consider a simple example of a board that has only two sockets; one to house a microcontroller and the other for a memory chip. On a usual day the memory socket is a battlefield for competing memory chips. An ‘alien attack’ happens when microcontroller suppliers come out with next-gen products that have integrated memory. Things outside of the semiconductor world – devices like cameras and digital calculators – have faced such an attack from smart-phones.

When faced with an ‘alien attack’ it is difficult to confront head to head. The crisis may not only involve price. For a customer using a single chip instead of two this makes sense for a whole range of reasons: cost savings, board space savings, ease of design etc.

The wise thing to do at that point would be to prepare a “flight” plan. “Flight” may not mean shutting down shop. It could mean scouting for new applications where the ‘alien’ threat is not applicable. As an example let’s consider the case of SRAMs. For standalone SRAM chips, a key market was Computer Cache at one point in time. In the mid-90s, Intel’s decision to integrate cache memory to the processor knocked out all the standalone SRAM makers from the PC market. In the years to come most of the SRAM makers quit the business, but some stayed on until today catering to other applications.

While defending an ‘alien attack’ is extremely difficult, making such an attack isn’t easy either. Especially, if the attacker isn’t as big as Intel. In the case mentioned above, what favored Intel primarily was their dominant position in the processor market. Years later when Qualcomm’s application processor attacked and killed the LTE socket in smart-phones a similar leverage proved useful.  When the ‘alien attack’ is planned by the company without such a leverage it might be complicated with a risk of failure.

In the case of my employer, Cypress Semiconductor, planning an ‘alien attack’ wasn’t too successful. When Cypress launched PSoC (Programmable System on Chip) the intent was to replace multiple sockets on the board including the microcontroller. PSoC’s value proposition was two-fold:

  • It allowed customers to reduce BOM (Bill of Material) cost by integrating multiple components to one chip
  • PSoC’s programmability made the design process simpler and quicker

While PSoC did win a few battles, the overall objective of replacing MCUs remained unfulfilled. One major roadblock in fact was novelty – traditional designers felt it was too different from what they were accustomed to using.

PSoC struck gold elsewhere. When the world began to move from mechanical buttons to Capacitive touch buttons, and from resistive touch-screens to capacitive touch-screens, Capsense and Truetouch (PSoC derivative chips) became a popular choice. The change in user interface format was in fact a new world application which made customers look beyond traditional chips. Thus for PSoC, the war became a ‘war for the new world’ rather than an ‘alien attack’.

Conclusion

In conclusion I would like to recapitulate the three fundamentals of fighting a socket war:

  • Know your war type
  • Know your terrain
  • Know your position

Having a clear answer on all of the above is essential to not just defending a socket, but to do so aggressively. In the Semiconductor world, fighting socket wars is the way of life. Not identifying the true nature of a war leads to an inevitable defeat. The truth is that some companies consistently fight it out better than the rest, and they are ones who survive in the long run. For the rest – shutting shop, declaring bankruptcy or getting acquired become the end of the road.

 

 

Posted in: ,
Tagged:

About The Author

Anirban Sengupta headshot
Anirban is a core-team member at Lifkart (an Early stage Indian Construction Start-up). Prior to the current gig he worked for about 5 years as a pricing manager at Cypress Semiconductor. He holds a BE in Electrical Engineering from National Institute of Technology , India and an MBA in Marketing from Symbiosis Centre for Management and Human Resource Development (SCMHRD), Pune, India.