Issue link: https://iconnect007.uberflip.com/i/929192
20 DESIGN007 MAGAZINE I JANUARY 2018 "Well, we want to use Gerber because it's unintelligent." Gerber wasn't made as a secu- rity method protocol for passing information. It was just what was around at the time in the late '80s to get something from the computer out into a physical board. But there is a para- noia about data being stolen, because it's a real- ity now that we're using more Asian resources. They have a piracy problem there, and not just with designs, but even, I'm sure every one of the people here today have had their software ripped off in Asia somewhere, us included. It's probably well-warranted, but it gets to the point where it starts to overshadow develop- mental stuff. And Martyn's right, we've had the same issue with customers. You've got to kind of guess what the problem is based on a description, because they're not going to show you what it really is, and it puts you at a loss with just trial and error basically. Am I solving the right thing for this guy? Shaughnessy: But do you all support ODB++ and IPC-2581? I know Cadence and Down- Stream do. Martyn, what do you think about IPC-2581 and ODB? Gaudion: We support 2581, but because most of our clients are pre-layout, then they're using our tools before there's too much data avail- able. The data we can put in 2581 is thin by nature, but we can. We have both links directly into the tools and via 2581, so there's two dif- ferent routes in. But there's not, from our per- spective, that much of it available to feed in, other than the stack-up information into 2581. Shah: Andy, one of the things that you can- not use earlier in the process is ODB++ for exchanging the stack-up data that Martyn talked about. Only 2581 allows you to do that. And I know that the adoption of 2581 is really gaining momentum. At the consortium update session at PCB West, Axiom Electronics talked about how they have standardized on 2581 and if anybody gives them Gerbers, they have to pay more, because it's an inefficient data format. Aegis supports 2581, so a lot of fabri- cation companies out there and manufacturing companies out there support it. Shaughnessy: How about printed electronics? Is printed electronics making a dent in your customers' jobs? Are you all seeing any of that? Shah: We do get regular inquiries from custom- ers whether we support that or not. We sup- port some outputs that support the printers. But you know, there's no industry standard, so there are multiple different languages that are consumed by the 3D printers. But it's an emerging industry right now, and people are prototyping with smaller boards, simpler tech- nologies and so on. Many companies are using 3D printers for developing prototypes. Shaughnessy: But you're not going to print 100,000 of them. Shah: Yeah, exactly. They aren't manufacturing in high volume at this point. Shaughnessy: It's a very cheap way to do cer- tain prototypes. Almeida: Right. That was the same experience almost everybody had when everyone was talking about "system on a chip." It was like, "Oh, well, you can put the whole thing on a chip, it's going to make PCBs obsolete." Well, you've got to walk that all the way through, what are you going to put the chip on? It's got to go somewhere! And what's interesting is we've seen for a while electronics shrink- ing down, and they've gotten to a point where they can only shrink so far without some level of voice recognition. You can only take a cell phone and make it so small because at some point, your fingers aren't going to work on the device. There's a human limitation that we have with miniaturization, and what you're Am I solving the right thing for this guy?