Will the real ESL market stand up? A nod to Gary and Lucio

Hey folks, John Cooley just posted a new piece on Deepchip.com regarding market research firm Gartner folding Gary's Smith's EDA analyst group. It's not surprising to note that most folks in Cooley's piece say they will miss Gary's group. There are a few that say essentially good riddance and that Gary's been too focused on ESL, which has yet to really step up as a formidable EDA market—in fact, with the recent news of Mentor acquiring Summit, it may be consolidating.

I've always took it that Gary was a big advocate of ESL as a new avenue of growth for EDA but I think he gave considerable play to the DFM space as well as traditional EDA market segments.

ESL's a rough market though—There aren't that many architects out there and if you're trying to sell ESL tools to software developers, they will freak out when you show them the price tag for EDA tools. Software and embedded designers are used to low to no cost software, so seemingly to be effective, EDA vendors have to be willing to sell a whole lot of software at a low price point and build up a massive support staff to support the many users or they have to focus on selling really expensive tools to the hundreds of architects out there to make R&D worthwhile.

But there is also a third way to do it and that's basically sell ESL as a giant bundled service: software for hardware software co-design, virtual platform development and model building.

That's what Synopsys is going to do anyway and it's all built around the Virtio acquisition. The company builds virtual platforms for TI, Intel and Freescale and Synopsys plans to add more to the channel.

Of course one of the biggest problems with virtual platforms is building the models. Thus far models are hard to make, even if they are in a standard language like SystemC, and putting them together to develop a virtual product is even harder for the average Joe and Jane designer. It's a curse for users and the traditional tools sales model but it may be a blessing for vendors that offer the whole ESL enchilada, like Synopsys and CoWare and seemingly Mentor with the (Summit acquisition) even if the vendor has a proprietary language for modeling. (And where Virtio has a foothold in the cell market, Alain Labat's company VaST has similarly formidable following in automotive—auto makers were gushing about them in my first feature of the year.) ESL indeed may be a razor sales model: selling the razor for cheap and making the real money on the blades. So perhaps the promise of ESL may not evolve as a traditional tools market: rather, it may come to fruition as a services gig much in the same way PDF Solutions sells its services to fabs and foundries. PDF, too, started as a traditional EDA sales gig, but the tools were too hard to use so PDF had to ship PhDs with the tools to run them. Making lemonade out of lemons—PDF hired an army of PhDs to be AEs. It worked. Now you have a good company in PDF solving a bunch or problems for fabs. See Lucio (Lanza), EDA is evolving the way you said it should—they just don't want to give you credit for it.