The cycle might take longer, Peter, but we will inevitably hit a standardisation phase. Look at how ARM and Co. are developing standards for server solutions, Android is becoming the dominant Mobile OS etc. Certainly it won't be like before where one technology or a few are standardised but it will be about system standardisation with a variety of nodes with standard interfaces.
I read this more as Synopsys aggregating EDA technology and segmenting it from its other business. That makes sense to me. As far as re-aggregating EDA within a large chip vendor, it's an interesting idea, but I don't believe we are close to that point. There are only a handful of chip companies that could realistically buy Synopsys, and I don't believe there is any real motivation for them to do so. When it gets to the point that only one or two firms are scaling to new nodes, then they might be forced to bring EDA in house.
Makimoto's wave gained some credit ability for a while. But as time goes by it seems to me the wave has failed to repeat....the key transition was from building block components that were never application specific... To soc era where all components are app specific.
Not just soc components.
Even DRAM has become an app specific PC memory these days
As you say Peter there are fewer companies willing or able to buy state-of-the-art million dollar EDA tools..hence your revenue doesn't grow...you have to do something about it...consolidating R&D of various product lines makes sense, I think EDA companies traditionally have spent too much on R&D % wise...Kris
Blog Doing Math in FPGAs Tom Burke 23 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...