What's holding back ESL?
June 03, 2013
Earlier this evening, Gary Smith held his annual night-before-DAC talks at the Austin Convention Center . During his presentation, Gary laid out his vision of the EDA industry over the next 10-15 years. There were many interesting points in the presentation (see Richard Goering's writeup for more details). One comment in particular stood out:
"Not as many RTL guys are moving up into ESL. They're really not capable of grasping the software end of it. So what we're getting is a lot of brand new engineers, young guys... They're young guys, they can do the software, they can do the hardware..."
My question to Gary during the session was to ask whether or not he was suggesting that ESL was being held back by senior design engineers, and that the migration to ESL might simply have to wait until a new crop of engineers with a background that included both hardware and software might be required to fully make the move to the newer design style.
Gary pointed out that when synthesis started making inroads into digital design, only 40% of blocks were being synthesized, and that number increased over time as newly designed blocks replaced blocks that were being reused. He feels the same trend is likely to occur here.
What do you think? Is ESL simply going to have to wait until there start to be a critical mass of IP blocks written in that style before it really takes off? Or is the issue simply one of training, where a differently-trained group of engineers will eventually be required to really get the ball rolling?
What, if anything, is holding back the adoption of ESL?