Six months ago, I predicted that emulation would be a bright spot within the EDA industry in 2008. Who could have guessed back then that our little world would be upended with the announcement of an astonishing attempt to merge two industry giants, both of whom have emulation products?
My optimism remains intact, however. First, the poor reputation of older generations of hardware emulator platforms has been put to rest forever. New emulators run faster; provide larger design capacities; are easier to use; have smaller footprints; and are significantly less expensive. They are also proven productivity boosters. Today, design teams are making strategic purchase decisions with emulation as a key component.
Emulation is clearly an enabling technology for system-on-a-chip (SoC) design and complexities associated with it as hardware and software elements need to be designed simultaneously. The software content in these designs requires fast validation tools, such as emulation. Emulation players, accustomed to hearing this from design teams they are in contact with, listened constantly to this refrain during the recent Design Automation Conference, as well. We’ve seen interest in emulation from a broad spectrum of applications—from processor companies to digital camera, consumer, and wireless markets. They all share one common necessity; they require billions of verification cycles to move meaningful data around. But other sectors such as automotive and aerospace are looking toward emulation solutions as well.
Further, we’ve noticed a change in the deployment of the new breed of emulators in the development cycle, i.e., a migration from traditional in-circuit emulation (ICE) to transaction-level co-emulation, without compromising the execution speed. Made possible by the availability of transaction-level interfaces and by new behavioral compilers that ease the creation of transactors for emulation systems, the new approach is moving the emulator upstream in the design cycle, a move that has opened the door to new applications such as ESL emulation.
But what about the battle of the EDA giants? The hostile attempt by Cadence to take over Mentor Graphics has left the EDA industry reeling and the emulation space in a tizzy because they are number one and number three, respectively. This leaves both providers and design teams with mixed feelings about this deal and trying to envisage the outcome. In the emulation space, it means that one of the two major players would be eliminated and disappear from the market, opening the field to newcomers large and small. For sure, this acquisition, if consummated, will be challenging for Cadence, not to mention the heavy financial burden to its bottom line. It will also be a painful experience for Mentor Graphics’ employees.
A longer-term, more positive view is that a merger of this magnitude could be an opportunity for the EDA industry to grow at a faster pace once again. It may offer EDA startups—especially in the emulation space––a path to being more successful and innovation might have more chances to prevail. After all, entrepreneurism is part of the EDA culture and a part that we do not want to lose.
My prediction from January continues to hold true: The emulation space has plenty going for it. It’s an exciting and growing part of EDA and it is shaping up to be a stellar year, with or without the merger of EDA giants. Check back in with me in 2009 to see if I’m right.