Concertal Systems SDA: Holy Grail, the Musical

Last year, Texas-based Concertal Systems came to DAC, but only as an observer. This year, the 15-month-old startup actually exhibited at the conference and, per reports, gave a pitch-perfect debut of their newly announced offerings: A suite of web-based tools – Audition, Orchestrate, Rehearse – which the company says effectively solves many of the problems associated with SDA, system design automation.

Per Concertal Founder and CEO Bob Ledzius, the automation of modern system design is completely dependent on the ability to seek out, vet, and integrate IP. But that process has been stalled, until now.

“A lot of people have been promising these capabilities,” Ledzius said during our recent phone call, “but we’re actually delivering on the promise with our SDA methodology.

“A lot of people have been doing great stuff to solve the incremental problems around system design, but we started by asking the right questions, not just the incremental ones. We took a step back and said, if you want the ability to develop designs in minutes – not months – then you need to develop a system for high-level IP reuse.

“Starting from this point of view, we came up with specific conclusions and solved the problem of IP verification and integration.”

Asked how the newly announced Concertal solutions were received this year in Austin, Ledzius said, “Nobody knew who we were coming in, but DAC was a big success for us. It was great.

“We gave a number of R&D demonstrations of our capabilities, and had people in attendance who then brought their colleagues back to see the demo as well. It was exciting for us to see the reaction.”

“Last year, the ink on our business cards was still wet when we were at DAC,” he noted. “We spent that time scoping out the technology, getting an understanding of the problems the industry collectively needed solved.

“This year, we brought solutions to those problems to DAC, and pretty much hit the mark.”

Ledzius outlined the capabilities of the tools in the musically monikered Concertal suite: “Audition is a web-based tool the users goes to where they can specify the IP they need. Our system then accepts that IP and configures it by building a wrapper around it.”

“Orchestrate is a web-based tool for someone who wants to design a new chip,” he continued. “You can go to Orchestrate and begin adding the functional blocks you want to add to your design. You don’t have to worry about how the IP is integrating together, because we represent all of those relationships in a data-base view. You just select the features and push the Build button.

“We pull all of the IP from the data-base management system, and build wrappers around the IP to support how it will integrate with the other IP. We generate a highly flexible fabric that’s totally user configurable. You don’t need to be an IP designer to go from concept to fully synthesizable design code in minutes.”

Regarding the third tool in the suite, Ledzius said, “Rehearse is a TCL-interactive script environment, also web-based, with heavy simulation and scripting support running on the server-side.

“You can inherit self-check for the IP and begin configuring the IP, so that it works together as a system. As you build the verification test case, the configuration information is represented in a data-base form. We then use that data base to optimize the design.”

“Again,” Ledzius said, “all of our tools are web-based, which is essential.”

Impressed with the potential, I asked if Concertal Systems has developed the long-sought-after Holy Grail of system design automation.

“Yes,” Ledzius reiterated. “A lot of people have been promising these capabilities, but we are finally delivering on that promise.”

Where does the technology arise from, I asked, and where’s the team located?

“The technology represents a life-long effort on the part of all of us,” Ledzius replied. “We are a team that’s worked together successfully before.

“We had a reconfigurable computing company in the late 1990’s called QuickFlex. While developing that technology, we also began to solve the problems associated with bringing IP together quickly for design.

“Now we at Concertal Systems are focused on system design, with our team in several locations – some in Austin, some in Bangalore, and some in Dallas.”

“There’s great rapport across all locations,” Ledzius emphasized. “You won’t find a more dedicated team anywhere. Everyone at Concertal believes in what we are trying to do, and it shows when we are together.”

Given his enthusiasm, I asked about the company’s road map and funding.

Ledzius said, “So far we are self-funded, and are working on several bids for actual contracts for developing new chips. We have just started responding to RFQs.”

“So for the present,” I asked, “you will be working as design consultants?”

“We want to get paid for using our technology to deliver real chips,” Ledzius corrected. “In other words, to get paid for beta testing our tools.

“Once we’ve got a couple of design successes behind us, we’ll be engaging in a deeper way with some of these [client] companies, and then offering our technology in a broader sense to give options to the industry for a better way to work together.”

Espousing the potential, Ledzius said, “We are sort of an operating system on a chip – but more than that, we are a methodology which facilitates the operation of building IP. Others might have tools for this process, but they are not automated.”

Which brought up the larger theme for Ledzius: “Because everything we do is through cloud-based collaborative ecosystem software, the user only needs a browser to leverage design off of IP reuse.

“We leverage off of the IP market, but we’re not a one-stop shop for IP reuse. We facilitate the shopping for and managing of licenses, but we’re not competing with the IP developers. We are developing our technology through the cloud, and offering this capability to other companies, but are not competing with those companies.

“Instead, we are allowing all these different players to work together in a new way.”

It was all a little confusing to me, so I asked: “It seems like Concertal is sort of eSilicon meets Design & Reuse meets IPextreme meets eFabless, and maybe even some EDA companies. Are these guys your competition?”

“We don’t see the EDA companies as competition,” Ledzius began. “We still need the EDA tools to develop the IP and do the back-end flow, while we are sticking to the front end. But the way we are delivering the tools is simplifying IP design.”

“And yes,” Ledzius continued, “some of these companies you mention have done pieces of what we are doing, but none of them have done it with a holistic point of view. Yes, the way we deliver the technology does create overlap, but it’s a good overlap if you’re trying to bring these industries together.

“We all know that chips are more complex than they’ve ever been before, and growing beyond the ability of any single designer to comprehend the entire system behavior. Meanwhile, the IP market is the fastest growing part of the semiconductor industry, the driver being the complexity of the chips.

“At Concertal Systems, using our SDA methodology, we have the potential to build an ecosystem of EDA companies, semiconductor manufacturers, IP developers – either internal IP developers or third-party – and allow them all to come together to build chips in a different way than they’ve ever been done before.

“We are really very much differentiated from other companies in our SDA technology. We want to use SDA to bring the whole design industry closer together.”

“If you think about it,” Ledzius concluded, “if you can automate the process of front-end design, and how the IP comes together, it allows you to be much more playful in how you design.

“Today, you have to figure out the topology of your design ahead of time. But if you could automate the thing, you would be able to experiment and optimize without the concerns of whether you’ve got it all right from the very start.”

“Yes, we have a long way to go, and we’re just in our infancy, yet what we’ve been able to do so far is pretty incredible.”

Ledzius chuckled, “Somebody at DAC called us the Facebook of SoC design, and we agree with that.

“We have real technology, that’s scalable, that builds relationships, that addresses the complexity of chip design. We are not trying to displace other companies – we are trying to glue them together.”

**********************

Tags: Bob Ledzius, Concertal Systems, DAC 2017, QuickFlex, System Design Automation

This entry was posted on Thursday, July 20th, 2017 at 3:24 pm. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.