This time last year, we were in the middle of selling Altos Design Automation. Altos had listened to its customers and realized that memory and standard cell characterization had become a huge problem, thus an opportunity. Their first large customer was Virage, whose product at nodes below 45 nm required at least four times the amount of characterization of earlier nodes. Altos changed their emphasis from statistical static timing to characterization as a result.
This shift also gave us a somewhat unique vantage point to watch the havoc Magma was causing in the Spice market with FineSim. Synopsys up until 45 nm had an uncontested franchise in the memory and standard cell characterization market. Magma had captured a huge amount of sales directly from Synopsys in the memory characterization market segment, which is the largest and fastest growing part of the Spice market. From my perspective, Synopsys had to do something to stop the bleeding. Altos ending up at Cadence limited the field of opportunities and, at least from my perspective, it wasn't obvious that they could buy themselves out of their problem.
Boy, I was wrong. I wasn't thinking big enough. I was really surprised by Synopsys's acquisition of Magma. First there appeared to be so much bad blood between the companies I just never imagined them actually talking to each other about an acquisition.
Sometime in the second quarter of 2012, Magma will join Synopsys officially. This will change the EDA landscape significantly. EDA will grow as a result of the pricing discipline that will occur, first at Synopsys, due to their elimination of Magma competition in the digital, memory and mixed signal sectors, and followed by Cadence. This is probably not viewed by EDA's customers as a good thing, but for EDA it will help both top line and margin growth.
The result will be stronger stock prices for EDA's remaining public companies—assuming the overall market cooperates. With this enhanced currency, we'll see increased investment in products and an expansion of EDA's footprint. The obvious spots will be IP, SoC Realization and Bare Metal Software. I'd also expect to see Product Lifecycle Management and Design Management become important in 2012.
The Magma/Synopsys event requires a response from the market. Spice will continue to be hot, and I expect more business activity in this area as vendors move to expand their existing footprints. Cadence's analog/mixed signal/radio frequency (RF) Spectre franchise isn't immediately affected, but they will look for opportunities created by the absence of Magma as a second vendor to Synopsys in the memory and digital characterization space. Ansys will have the Apache acquisition integrated in the first half of 2012; I think they will naturally look to expand into an area adjacent to microwave, such as RF simulation. The same is true for National Instruments after the purchase of Applied Wave Research. A company such as Berkeley-Design Automation, whose product is a Spice transient and harmonic Spice simulator, would fit the bill as acquisition candidate here.
In a way it is non-intuitive that Spice, one of the first commercial EDA segments, would lead the current evolution of EDA. But then semiconductor design has always been about the transistor performance.
Jim Hogan, a 35-year veteran of the semiconductor design and manufacturing industry and a former executive at Cadence Design Systems, National Semiconductor and Phillips Semiconductors, among others, is now a private investor and managing partner of Vista Ventures LLC.