Issue link: https://iconnect007.uberflip.com/i/994883
JUNE 2018 I DESIGN007 MAGAZINE 13 tecture, and then at some point it ends up as PCBs. At the back end of the left-to-right pro- cess, we at Mentor have been leveraging the Valor technology to optimize what we call the lean NPI process to handle the transfer and the collaboration between the PCB design team and the manufacturing team. There's also top to bottom, where you have the multiple disci- plines. You've got hardware and inside hard- ware there's multiple PCBs that have packages on them, which have silicon on them, and then there is software, cabling, and maybe the networking on what's being communicated across that cabling, and the mechanical infra- structure. These are all different disciplines that, at some point, are being designed con- currently after the design has been partitioned up for those disciplines. Those guys need to collaborate more effi- ciently as well, top to bottom, and we've talked in the past about things like ECAD/ MCAD collaboration, right? How do you opti - mize those guys working across disciplines? The same thing happens between electronics and electrical. Is it in the wiring? How do I optimize? To put on my Siemens hat now, we talk about a continuous digital thread. The idea of a digi- tal thread is from that initial point of require- ments capture, where everything is continu- ously created and communicated in a digital form, eliminating any redundant efforts and optimizing the communication between steps. That's the Holy Grail. tecture was decided, they don't tend to do a lot of revisiting it in the form of trade-offs, in the form of any ECOs. When the design gets bigger, it becomes nearly impossible to use that form of collabo- ration, so that's an impact because of complex- ity. Also, because of that complexity and the density of designs, the performance of designs, you've got to think about trade-offs like form factor or performance where before maybe the box was big enough. You really didn't care what went where or you were operating down at much lower speeds, and so you didn't care about performance across boards, but as those changes happen, it becomes more critical to do design trade-offs. Regarding the second challenge, to address complexity, teams need to be able to com- municate or collaborate across multiple disci- plines. I was just talking about the box, and so I'm talking about the mechanical guy and the guy doing the PCB being able to collaborate. Maybe the high-speed engineers are collabo- rating with the PCB guys, and so this multi- discipline collaboration needs to be managed more efficiently, again, to enable them to do those trade-offs to avoid any redundant efforts. Because a lot of times when they pass data from one person to the next, they're doing it again through some manual means, maybe email. It may be a whiteboard. There are lots of ways to do it that are very inefficient, so somebody receives that data and they basically must redundantly design whatever that person communicated and, of course, that introduces lots of fun errors. We're really trying to find ways to break down what we talked about, black box design or silo design processes. There are silos really across the spectrum. If you think about a pro- cess from left to right, on the left you've got requirements and, by the way, left to right you could think about it in terms of a V diagram, too, if you wanted. On the left, though, you've got your basic requirements. Those market- ing people they come in and say, "Here's what we want to build." It has requirements, and then somebody takes those requirements and they progressively refine it down into an archi- Figure 2: Optimizing electronic systems design with a continuous digital thread from requirements through manufacturing, and across multiple domains.