I and the rest of the Cadence C-to-Silicon Compiler (CtoS) team were thrilled yesterday to learn that EDN selected C-to-Silicon Compiler as one of the "Hot 100 Electronic Products of 2008".
How about that, John Cooley!! www.deepchip.com.
Why does the launch of CtoS in 2008 represent a significant EDA industry milestone? Very simple.
HLS has been the "holy grail" of EDA for 15+ years, and with CtoS, we *really* think we found it.
When I first joined the CtoS team (on the marketing side -- I don't even pretend to be on the same level as the geniuses who actually built the tool) I was amazed at the list of architects/technical advisors. It read like a "Who's Who" of EDA, with names like David Knapp, Joe Hutt, Andreas Kuehlmann, Mike Meyer, Yosinori Watanabe, Luciano Lavagno, Alex Kondtrayev, and many others. I've been in EDA for over 12 years now, at both Cadence and Synopsys, and I don't remember there ever being a case of more "total brainpower" assembled together on a single project. Their mission was simple: make designers 10x more productive by building a high-level synthesis tool getting better-than-average-human QoR for *any* type of design.
Why 10x more productive? A higher level of abstraction means designing more gates/day, with fewer lines of code, fewer bugs, faster simulation/verification, and more reusability. This year we've been working closely with a handful of customers and so far the results are looking even better than we initially expected.
I'd be a liar if I told you we haven't run into lots of "speedbumps" along the way: beating humans at any type of design is not trivial (the universe of digital designs is pretty big!), helping designers learn to think/work in new ways is a stretch, dovetailing into the rest of the EDA tool-chain is complicated, so it's no surprise that our early CtoS customers are "holding our feet to the fire". But the awesome part is they're not doing it out of disappointment, or frustration...they're doing it because they're excited, as excited as we are.
That's why I think with CtoS, the time for high-level synthesis has finally "arrived".
Watch this space. In future blogs I'll keep updating you on our journey.