Fitzpatrick on SystemVerilog
SVUG and DVClub - March 20, 21, and 22

DVCon Cadence Panel and a Guest Author!

Terry Lawell, Verilab's VP of Business Development, joins us today as a guest writer describing the Cadence Design Systems lunch panel from DVCon a few weeks back.  Why the delay?  I've been prepping for a presentation that I gave yesterday afternoon at the Institute for System Level Integration in Livingston, Scotland, and as a result had to push my publishing schedule out a bit.  Regardless, here's Terry's take on the panel discussion.

One of the interesting events of the conference was the luncheon panel hosted and selected by Cadence. Mike Stellfox from Cadence was the moderator for the panel. The panel consisted of Jon Michelson, Claude Cloutier of XtremeEDA, Kevin Jones of Rambus, Steve Jorgenson of HP, and Brett Lammers of ST Microelectronics. The topic of the panel discussion was 'What is most important in verification, the language, the methodology, or somewhere in between'. Though the panel consisted mostly of people with 'e' experience, they had varying opinions on what was important. I think one of the most important and interesting things about the panel discussion was, though the panelist had different views on what was most important between language and methodology they all agreed that there was a third part to the equation that hiring smart people was essential to a successful verification flow. Once the notion of smart people entered the discussion, language and methodology tended to take a back seat. I think one of the issues around this topic that was not discussed was the idea of hiring smart people to develop the methodology in such a way that engineers with less capability could easily employ the methodology.

Another interesting topic that sparked significant discussion was the concept of dividing lines between design and verification engineers. Some of the panelist said they have a process by which they rotate designers and verification guys on a regular basis and this develops an appreciation of both disciplines, improving the skills of both. Most of the other panelist believed that the two disciplines required an entirely different thought process to effectively accomplish their tasks and believed the dividing line was necessary.

In summary, the panel basically agreed that the order of importance was:

  1. Hiring smart people
  2. Methodology
  3. Language