Where Have All the IP Vendors Gone? Part 3: Market Consolidation

We have answered the first two questions on Market Growth and Market Maturation. The third question relates to Market Consolidation and its impact on the long term health of the 3^rd Party IP market.

Does the fact that we have recently seen several ‘large’ IP companies acquired by other ‘larger’ IP companies mean the long anticipated consolidation in the industry has gotten underway.

In order to answer this question it might be a good idea to consider what conditions contribute to, or force, a market consolidation in the first place:

1. Worsening conditions in a specific end market or in the overall economy
2. A drastic change in the direction of technology and a lack of internal resources to keep up
3. Poor financial performance by a company in their chosen end markets making them vulnerable to a takeover by a competitor
4. No access to capital for business expansion
5. Remove competition from the market

In the case of the 3^rd Party IP market, do we see any of these conditions occurring across large numbers of companies or across a large number of end markets or applications which would prompt deteriorating financial conditions in any of the companies that were recently acquired – MIPS, Tensilica, Cosmic Circuits and now Evatronix? Or looking 2-3 years further back, ARC, Virage Logic and Denali? Given all the industry has been through in the last 4-5 years you could say that the financial health of any of these companies could have been better at any one given point in time, but were any of them in danger of failing, thus became targets for acquisition? The answer is no.

There are several factors that govern what everyone would consider to be a market consolidation. One metric is the number of companies in the market at any one time – is it going up or going down. Today there are approximately 350 – 400 IP companies in the market and the number fluctuates for any of several reasons. This by itself does not signal the overall trend one way or another. However, if we saw multiple IP companies abandoning the market, then you could reasonably assume something was happening outside the normal course of events. We do not see that happening today. If anything, the reverse is true with new companies entering the space. True, they are not household names yet, but give them some time to get established and then see what happens.

So if all this is true, what is really going on?

If you look at the companies that have been on an acquisition path in the last few years and what they have acquired you would see the following.

• MIPS: acquired Chip Idea for Analog – then sold off same to Synopsys
• Virage Logic: acquired ARC for CPU cores and audio IP
• Virage Logic: acquired the IP group from NXP yielding Interconnect and Interface IP along with several other product types
• Synopsys: acquired Virage Logic
• Cadence: acquired Denali for memory IP
• Imagination Tech: acquired MIPS for CPU IP
• Cadence: acquired Tensilica for CPU and DSP IP, along with IP Subsystem product
• Cadence acquired Cosmic Circuits for Analog and Interface IP (deal not yet completed)
• Cadence: acquired Evatronix for controller cores along with Interface and some Analog IP
• You could also add the recent agreement between Sonics and ARM whereby ARM has licensed the Interconnect IP patent portfolio from Sonic

In all these cases the acquiring company sought out and acquired IP that they either lacked or that reinforced what they already had in house. More recently, the acquisitions by Imagination Tech, Synopsys and Cadence filled in their existing portfolios with IP that is highly differentiated and provides expanded computational resources for applications that may require it. In none of these cases did the acquiring company compete directly in the areas the acquired company participated in and so did not present a great threat to the acquiring company in the market. There clearly is something else at work here.

To Semico these moves signal an increasing desire to build more comprehensive and complex products using the acquired IP in combination with the IP already in the possession of the acquiring company.

In other words, they are going to build IP Subsystem products. In the cases of Synopsys and Cadence, these companies already were IP Subsystem vendors. Now they will be able to cerate even more powerful and diversified subsystems using the new acquisitions. In the case of Imagination Tech they will be able to add a world class CPU core to their existing offerings and improve the performance of these products with an eye on creating even better products in the near future. These future products could include IP Subsystems of different types because they now have better computational resources to work with. More resources = more room to differentiate future offerings and the ability to make them more powerful; something the market is always looking for.

To Semico this is really what is going on in the market today. It is not market consolidation or market maturation or market slowdown, it is the next stage in the evolution of the IP market – the IP Subsystem phase. In this phase IP companies are looking to create products that are more comprehensive in the solution offered and are self contained and not just small parts of a larger product to be assembled later. While it is true that IP Subsystems are not necessarily standalone products in their own right, they do offer the designer the ability to design with a larger piece of the puzzle.

Look at it this way: instead of the designer needing to identify, acquire and aggregate 200 puzzle pieces that will become the system-level functions needed for the solution, he now can acquire the system-level solution itself and infuse it into his design. The more subsystem products used, the greater the reduction in the level of effort expended to create these functions and time and resources are freed up to do other things that are more beneficial to the designers and their companies.

In the discussion about IP Subsystems, there is an additional important area to point out and that is Software. IP users today are asking their IP vendors to provide more software with the IP they are licensing in the form of drivers, APIs and, in come cases, the applications that will run on the IP block. Rising software development costs are prompting these requests. Software design costs at 28nm were just about even with silicon design costs and now represent a slight majority of total design costs at 20nm. There is every indication software design costs will increase at each new node going forward. What better way to reduce the integration effort than to provide some of the software needed for the silicon solution and include it with the IP Subsystem being delivered to the designer? It makes a lot of sense to do so and allows for a great deal of differentiation in the end solution. This is a path that many IP vendors are starting to take.

It is our belief that the industry will see additional acquisitions over the coming months, and they will be mostly selective in nature. Some portion of these acquisitions will be identified as companies that hold pieces necessary to fulfill the requirements to build one type of subsystem or another or that help the acquiring company realize the goal of building a subsystem product.

If IP vendors really thought the market was slowing or maturing, they would not pursue the acquisition path since it will require time, money, resources and a plan to execute successfully. In a contracting market, these are precious commodities and not to be wasted on acquiring additional types of IP in the face of worsening market conditions. No, it is much more likely that each acquisition provides a missing element to the acquiring company that allows them to do something they couldn’t do before. In this instance it is building IP Subsystems to lay the foundation for even better silicon solutions and higher market revenues in the future.