Cadence announced that they had acquired Denali for $330M today. They reckon that it will be accretive this year. Since the beginning of the year there has been a sort of land grab going on in the system and IP spaces. There are rumors around that at least one of the high-level synthesis companies is about to be acquired too.
It is going to be interesting to see just how this all plays out. Magma isn’t playing this game, they are sticking to IC design for now.
Mentor is the market leader in high-level synthesis (HLS) with their Catapult product, and they have had an embedded software and RTOS business for many years. However, they have no virtual platform technology. Mentor also has a strong PCB offering, another important part of design once you get up off the raw silicon.
Cadence has a relatively new HLS product, CtoSilicon, they have a partnership with Wind River (Virtutech) around Simics for virtual platform (although this is brand new so too soon to tell how it will work out) and have had no IP portfolio to speak of until today’s Denali acquisition. I’m guessing that Synopsys would have been interested in Denali but didn’t want to pay that much. Cadence also has a strong PCB design offering.
Synopsys has no HLS product, although rumored to be in discussion to acquire one. They have Virtio, VaST and CoWare in the virtual platform space. And they have the strongest IP portfolio of any of the big companies. However, they have no offering in the PCB space.
It is going to be interesting to see how this plays out. The problem for all the EDA companies is that as you move up to higher levels of design, then “almost all” designs are FPGA-based and do not involve designing a special ASIC/SoC/ASSP. This does not play to these companies strengths. For the FPGA/embedded market their channel is simply too expensive and doesn’t call on the right accounts. When I was at Cadence, their system design tools were in a separate organization, the Alta Group, which had its own sales team. The business was growing nicely. Cadence decided to optimize expenses and roll the Alta Group back into Cadence proper and take away the separate sales team. From that point on their products pretty much went into decline and eventually Cadence transferred them to CoWare in a deal involving transferring the team, getting royalties, taking some stock and maybe even some real money (so amusingly those CoWare businesses are now part of Synopsys and presumably Cadence got some money out of the deal for their share of CoWare).
It will be interesting to see if any of the big 3 EDA companies bites the bullet and puts its system strategy into a separate organization with its own channel and, perhaps, logo. I have a feeling that it will be a necessary condition for success once a critical mass of products exists. As I said about Cadence’s EDA360 announcement, real success will be if Cadence is selling significant product to companies which are not doing IC design. The same goes for Mentor and Synopsys.