Pros and cons of creating a combined database and user interface to bridge these disciplines.
For many years, the design tool industry has entertained the idea of combining both electrical and mechanical design into a single user experience, with a single database as a foundation.
Major tool vendors, at least on the electrical side, have taken the matter seriously and confirm that activities towards a single flow have been considered, particularly as the EDA industry has long searched for opportunities in adjacent markets. Considering that the mechanical design tool market is roughly double that of EDA, it’s easy to see why it is a good candidate to cozy up to.
“That’s been where our largest share of growth has been,” said Rhines, chairman and CEO of Mentor Graphics. “We don’t just compete with Cadence and Synopsys. We compete with Ansys, Dassault, Autodesk, and that’s going to be true of our whole industry. But there is a unique constraint here, which is that it’s still electronic or electrical design. For years, I used to hear, ‘Why don’t they integrate the mechanical design with the electrical design?’ After all, these things need to work together, but the fact is that engineers who design electronics have a way of looking at the world. They have an interface they want to use, and so when somebody introduces one that handles both the mechanical and the electrical, nobody adopts it.”
When Mentor jumped into wiring harness design software, many EDA companies questioned the move. “After all, you have to predict that this wire bundle will fit through the hole in the doorframe, and a car is a three dimensional vehicle,” said Rhines. “It’s not a two dimensional layer in an integrated circuit. How do you do that? The answer is that you give the electrical designer his or her interface that they prefer, and you make your software go and grab the piece of information out of the mechanical design software that gives you the dimension for that hole in the doorframe. Then you verify the electrical design with the hardware design. This is one of the reasons it has taken Mentor so many years to become the leading supplier of electrical wire harness software. But the fact is that mechanical design software companies still do mechanical software design, and they recognize that it has got to accommodate what the electrical designer does. Electrical design software companies do electrical design, but we do gain access to and utilize the data in the mechanical database.”
The people using these tools come from different backgrounds, though, and there is a push among EDA companies to bridge the different worlds by understanding how each side approaches a problem and making tools approachable and usable for all of them.
“Both electrical and mechanical engineers have a certain domain experience and we shouldn’t make them learn the details,” said Anush Mohandass, vice president of marketing and business development at NetSpeed Systems. “What we try to do is simplify the problem of cache coherency. Why is it a tough problem? Because the software guys try to solve it, and they are terrible at it. They don’t understand how hardware works. A better solution is to hide these difficulties in the hardware and make software engineers do what they are good at. This is very similar to mechanical and electrical design in the area of ADAS, where there are automated cars or robotics. If you make the mechanical engineers understand the intricacies of electrical design, you’ll fail right there. You have to set it up so that the mechanical engineers do what they are good at, and electrical engineers do what they are good at. Both of them should understand what the end system is, but at the same time not get to a point where they need to understand every intricate detail. Combining them, my worry is that instead of getting the best of both worlds, we get the worst of both worlds. Each side is mired in their own world, and trying to understand what the other discipline is trying to do, and getting lost in the details.”
So while the convergence of these two worlds is inevitable, the merging of the mechanical and electrical disciplines is not.
“We’re getting to a point where the separation between the mechanics and the electrical is now blending,” said Sundari Mitra, CEO of NetSpeed. “It needs to get combined, but how it will be combined into a single user interface and database is a challenge for the industry to confront rather than shy away from. Mechanical engineers are going to want to work at it at a much higher level than the electrical engineers want to. There can be a separation and it could be part of the same interface.”
Mitra stressed that the whole system needs to be presented as one for a successful outcome because the mechanical and electrical aspects are co-mingling. “Keeping them and treating them as two unique problems is going to pose difficulties, but the challenge is that we need to come up with a database structure. We need to come up with methodologies that allow the right level of abstraction so that you don’t end up getting the worst of both.”
In the area of coherency, she explained that the ‘worst’ could be that a software engineer has to deal with all the hardware issues. “Not having a background in that, and being forced to design something—that is the worst of both. Or there is a hardware engineer whose whole perspective of coherency is, ‘This is the amount of memory that I need, this is the bandwidth that I need, let me design a system that doesn’t give them exposure to this level of detail.’ They are really going to mess that one up.”
Software vs. hardware
Bridging software and hardware is a huge problem, but it’s also one that the EDA industry can no longer ignore.
“The way the world is changing, products have much more of a software focus, said Simon Davidmann, CEO of Imperas. “And with software a $3 billion market, should EDA be related to embedded software? Absolutely, in the same way that EDA came from pushing polygons around—CAD, then schematic capture, RTL, verification, IP. And and IP is just a better way of designing blocks. You buy them rather than build them from scratch, so it’s part of the ecosystem that has evolved. Should it be part of our industry? Should we be working closely with it? I don’t think the IP industry is a solved problem in any way. It is still very hard to put a large chip together with all of the different components required. We haven’t solved all of those problems, and the industry has to evolve into adjacent industries because the customer problems are in those industries.”
Further, he pointed out, there’s no point in just having RTL design tools. “You need tools that allow you to bring in IP the same way you need to bring in software. Do you need to bring in mechanical? Clearly, there are lots of synergies between the challenges that are there—packaging, thermal, cabling. If you can package chips into 3D devices, you can get the memory much closer to the CPU, and you can speed things up. As such, from an architectural point of view you might want to do stacking dies, and that becomes a mechanical packaging technology, which is part of EDA but it’s not electronics. Is this branching out? It touches it already.”
Where does thermal fit?
One area of technology that appears to be bridging the electrical and mechanical world is thermal.
Is thermal a mechanical issue? “It’s definitely not electronic,” Davidmann said. “It’s product design, so we need to think of the challenges that product design has. Our job is to help people build products and get products out the door. Some EDA companies have tools to help build embedded GUIs. Is that an electronic product? Is it software? What is that? Mechanical is just the next bit of that. Should there be a close relationship? Absolutely there should be. Mechanical is about a $7 billion market. It’s a bigger business, a different business. Should they be better related? Absolutely, in the same way Mathworks is related to it, which is how it works and how the business flows.”
The opportunities in the automotive industry have also added credence to the argument for combining mechanical and electrical design, particularly where antenna design is concerned, according to Vic Kulkarni, senior vice president and general manager of the RTL Power Business Unit at Ansys.
In his mind, it is a question of timeline because connecting the dots is critical given the role electronics has in controlling mechanical parts now. Take Industry 4.0, for example, which can be used to create an electronic device to control a factory floor. “The robotic arm, for instance, can be controlled from an iPhone on a factory floor, and then you can change the recipe on the fly to create customized vehicles, customized products of whatever you are manufacturing—a multi-physics approach gives that advantage.”
Frank Schirrmeister, senior group director for product management in the System and Verification Group at Cadence, believes there is potential value in having electrical and mechanical in one database, similar to the way analog has been blended with digital. But the question is whether it should be defined in one database. “The database piece is less of an issue. It’s really a question of doing something in conjunction with it. Should mechanical and electrical design be combined into a single user interface and database? There’s an education issue. Nobody has a grasp of it all so nobody would be able to make sense of it.”
Just as with a construction project on a home or building, each area such as plumbing, electrical, and structural have specialists to do the work, Schirrmeister said. “There are reasons for the database to be connected, in a sense, because you want to make sure that the data is consistent from one database and you don’t make any mistakes when adding in components.”
Still, even with all of the advances in technology, there still is not a single interface and accompanying database for a combined mechanical-electrical flow. And Rhines maintains that is never going to happen.
The Trouble With MEMS
Severe price erosion is putting this whole sector under pressure at a time when demand is growing.
Evolving Thermal Landscape
How finite element analysis and other technologies are being used to reduce risk and uncertainty of thermal impacts in advanced packaging.