David Vye, Technical Editor, Microwave Journal:
Microwave Journal had the pleasure of speaking with Mr. Todd Cutler who is currently a Senior Manager with Agilent Technologies’ EEsof EDA Division, responsible for Marketing, Services, and the ESL Business. In his 28-year career, he helped to found the division, which began in 1986 as the computer aided engineering group within Hewlett Packard’s Test and Measurement group. Cutler later left H-P to be CEO of Eagleware Corporation, the EDA company that pioneered the affordable Genesys RF and microwave circuit and system design software platform. After seven years with Eagleware, Cutler helped broker the deal for Agilent’s acquisition of Eagleware in 2005. Here is an interview with Cutler where he shares his perspective on the recent integration of Agilent’s Momentum 3-D planar EM solver into the Genesys platform.
MWJ: Since the Genesys environment product already had a planar EM tool known as EMpower, why did Agilent decide to integrate Momentum?
TC: Empower is an accurate, capable simulator, but being grid-based, its range of application is limited. For example, when the spacing between lines is not a rational factor of the line width, the size of the grid must be small and the time to simulate grows dramatically. Momentum, being mesh-based overcomes those limitations. The net result is that our customers are able to solve more general structures much faster. This is a great example of how the combination of Agilent and Eagleware is providing new value to our mutual customers. What’s particularly good is that we integrated it with Genesys, keeping the easy, transparent circuit-EM simulation that Genesys is famous for. Our customers tell us that they want all the simulation tools they need in a single environment, and the superior technology in Momentum, when deeply integrated into Genesys, preserves ease of use while allowing designers to address new EM requirements.
MWJ: Is there a difference between Momentum and Momentum GX? Do ADS users and Genesys users have access to the same level of capability from the Momentum product?
TC: Like Genesys, Momentum GX is packaged to address the needs of self-supporting RF workgroups who don’t have access to a lot of resources, but still need to do quality work. From the perspective of the typical Genesys user, Momentum GX preserves the accuracy, co-simulation, meshing, loss, stack-up, and other features relative to the ADS version. Since ADS is aimed at a wider range of applications and technologies, it has additional distributed licensing and solver capabilities, as well as links into high-end IC design and multi-technology design flows that make less sense for Genesys.
MWJ: Can you give us an example of the types of problems that Momentum GX will be able to solve that would have been problematic before?
TC: Yes. Large and complex RF board layouts with non-orthogonal interconnects, as well as high-end filters and resonators with novel shapes, among others. The Momentum GX is much faster and has both full-wave and quasi-static solvers. This allows larger analyses for a given amount of memory, and therefore more complex structures. The Momentum GX full-wave solver also allows open boundary problems, unlike Empower, which relies on a method of lines solver within an enclosure.
MWJ: Sounds like the technology behind the Momentum EM simulator is significantly different and more powerful than that available with the previous solver. Does the integration of Momentum into the Genesys environment increase the price of the combined product or is the new EM capability an optional add-on?
TC: The price of Genesys remains unchanged at $4K (US) for a linear physical design environment, including layout and libraries, which is unprecedented for its set of capabilities. Momentum GX is a $10k (US) option, plus support.
MWJ: Is there an upgrade fee for existing Genesys users or is it a part of their standard maintenance upgrade?
TC: Momentum GX is a new simulator that can be added to any supported seat of Genesys. Empower owners have the option of adding the new Momentum GX at a reduced price.
MWJ: Who and what specific applications do you see most benefiting from the combination of these two products?
TC: People who want to miniaturize a board-level design, or extend dynamic range and sensitivity, or go higher in frequency, or bring passive designs in-house from a vendor, or move to a new technology… all these design tasks require accuracy within the RF EDA platform. Momentum GX enables better designs and reduces prototyping requirements by providing an additional level of physical design verification at the point of design.
MWJ: The Genesys product seems to target smaller companies that may not require the horsepower of ADS and have restrictive budgets for engineering tools. What are some of the features that these companies are sacrificing for cost considerations and how do you help companies assess their true needs?
TC: Sometimes the choice of design flow is an external business constraint. For example, one measure of the effectiveness of an EDA tool is the number of board turns saved in a year. If you are a smaller organization, you may not save enough money in board turns to invest in high levels of automation or enterprise-type tools, or the staffing that goes with those tools. Seamless, multi-technology design flows create long-term competitive advantages, and people do this with ADS all the time. However, sometimes the best business decision is to buy RF and CAD/Manufacturing tools that have the raw technology you need, and just live with less automation. Genesys is superb for doing traditional RF & microwave planar designs with a minimum loaded cost, which is great for start-ups, and now with local-language support, emerging economies.
MWJ: The Genesys product is well-known for its excellent synthesis capability. How well does the addition of Momentum into the Genesys product support a synthesis to verified layout design flow?
TC: Genesys has about ten such synthesis modules, and the ones that produce planar physical structures have been updated to interact with Momentum GX. For example, we’ve added a button on the M/Filter module to “prep for Momentum.” It does all the setup, including ports assignment, boundary conditions and so on. The designer just adds a Momentum GX analysis and presses “solve.” This illustrates the high level of integration we are aiming for an achieving, going beyond a “socket” approach.
MWJ: Wow, that sounds like an impressive amount of automation and time saving for the engineer. So the synthesis to EM simulation flow is very user friendly. How about the rest of the flow? While EM tools are used for design verification they are also becoming popular in actual design vis-à-vis circuit/EM co-simulation and parametric studies. It sounds like this integration does support circuit-driven optimization using results from Momentum? Is this the case? Since EM simulation generally takes longer than circuit simulation, are there any capabilities such as job queuing or distributed analysis available to manage multiple parametric analyses or speed up the overall simulation run-time.
TC: We’ve taken the first step, which is to invest in co-simulation, verification, and design for manufacturability, and will continue to invest in higher levels of design automation, especially with more application focus. Agilent is enabling deeper EM capabilities in all our design platforms, not just Genesys.
MWJ: On a related subject, does the EM simulator tie up the environment while it is running or is the engineer able to continue using the interface for other circuit set-up, simulation, post-processing or other operations?
TC: The UI is available, but gets fewer CPU cycles during an EM simulation. Some simulations run in seconds, so it’s not always such a trade-off.
MWJ: I’ve looked back at interviews from the time of the Eagleware-Elanix acquisition and saw you mention that a number of companies owned both products (Genesys and ADS). Is this still the case and does that indicate that both products have unique capabilities and advantages? If so what are some of the areas where these products do not overlap?
TC: Some of our larger customers do own both products. In 2006 we responded by making the unique Synthesis and Spectrasys features more directly accessible from within ADS. These capabilities add value as point-tools to any design flow, in fact. As a platform, Genesys serves the area of the RF design community that is less receptive to the value of the larger enterprise tools. Genesys allows people to succeed out of the box, with less overhead. It is easy to use, compact, affordable, and has enough performance and variety of simulators to get the job done. Agilent is in a great position to have platforms optimized to cover the wide spectrum of needs in the industry, and makes it easy to configure a solution that you won’t outgrow.
MWJ: From the product feature article in Microwave Journal it appears that the Momentum integration works directly with the system tool – Spectrasys. System simulators seem to be about top-down design whereas EM simulation is about design verification, parasitic analysis and passive component optimization. Does this combination of top-down and bottom-up design capability signal a new level of sophistication among smaller engineering firms?
TC: It’s important to have realistic component performance at the system architecture level so you don’t design a system that can’t be implemented. Unlike Excel, Matlab and other tools, Spectrasys accounts for a lot of analog effects and is very interactive. If poor filters or board parasitics allow harmonics and spurs to leak across your system, that is something you’d like to diagnose quickly, and avoid. Being a large test equipment provider, Agilent has the most comprehensive verification environments. Just like it’s a business decision to invest or not-invest in seamless design automation, seamless verification is also possible, but not always practical if you have a sporadic or high-mix project flow.
MWJ: The real world can certainly be more complex and challenging to design for when high frequency effects are taken into account. With this being the case, it seems as though engineers coming out of school will really need to know their design tools. Do you find this to be the true and are there any advantages to training on any specific EDA circuit tool.
TC: All of today’s EDA tools are growing in sophistication to keep pace with today’s increasingly complex applications. Many major Universities are addressing the need for EDA tools training through training partnerships that we offer. In terms of educational environments, Genesys itself is easy to learn and can be used more transparently in a compressed time frame.
MWJ: “Easy to learn” is an essential requirement as design tools evolve to address greater circuit complexity and new performance metrics in both time and frequency domains. How does the new EM simulation work with other Genesys components such as the frequency planner called WhatIF or the transient SPICE simulator known as Cayenne.
TC: Genesys is an integrated platform, so Momentum GX is visible to all other parts of the platform environment.
MWJ: Genesys targets (or is at least positioned as) a lower price/performance point product than ADS. It seems to be the natural evolution of products to add capability and thus increase price. How does Genesys maintain its price/performance position as new capabilities are added? Do you see the whole RF circuit market moving up the price/performance curve or does this market segment become more of a commodity?
TC: While the market as a whole shifts to accommodate the latest designs and performance requirements, a spectrum of need for EDA design tools will always exist, from individuals to groups as part of a large enterprise design. Genesys and ADS cover a wider spectrum together than either could do apart. We are investing in both to stay ahead of technology trends across the spectrum.
MWJ: Does an engineer or design group have a migration path for moving individual designs or entire design inventories from Genesys to ADS? Is it the company desire to introduce users to Genesys and then move them over to ADS as their needs and budgets grow?
TC: In working with Agilent, a company can start with Genesys and stay with us as their needs shift, because we have the spectrum of tools that can scale to meet a company’s growing needs. For example, last year’s investment in making Genesys tools more available to ADS users has helped to facilitate the use of both products and bridges the gap between them. But at the same time, Genesys is itself a fully-capable environment. We have thousands of customers who are perfectly happy with Genesys, and there is no pressure to forcibly migrate them to ADS. That is exactly the point of adding important EM capability within the Genesys platform. We hope this is all a compelling reason to do business with Agilent.
MWJ: You’re right; I can see the value in offering a wide portfolio of products to a market that has as many diverse needs as ours. While adding Momentum GX adds a necessary level of capability and accuracy to the Genesys product, a number of Eagleware/Elanix users expressed concern that the Genesys and ADS products would eventually be merged. At the time Agilent acknowledged that there would be some “cross-pollination” between products, which seems to be exemplified by this announcement. Has the merging of these technologies in this instance been accepted as a good move or does it feed into those earlier fears?
TC: Well, I can say that we’ve shown that any early customer fears are unfounded. This is our sixth release of Genesys since the acquisition, and both sets of users have benefited from the cross-pollination. As I mentioned earlier, we believe that we are now uniquely suited to meet the needs of a broad spectrum of designers.
MWJ: Are there any plans to introduce 3D EM co-simulation into Genesys using the new Agilent 3D solver or through your partnership with CST?
TC: We are investing heavily in EM technology right now, and are open to all the possibilities that make technological sense.
MWJ: The acquisition of Eagleware/Elanix along with flagship product ADS, and RFIC simulator GoldenGate seemed to be a corporate decision to have a broad portfolio of design tools covering the range of RF design from RFIC and MMICs to RF boards and from small design house to large enterprise. From a business perspective has the strategy worked as planned? Has the Eagleware/Elanix community been happy with the direction of the product and the resources that Agilent brings to the table?
TC: This is, in fact, the plan. Some members of the Eagleware community were apprehensive early on, but now that we are in our 6th release in 2 years, customers see the results of our ongoing investment in the Eagleware product line. We’ve seen our customer satisfaction rates rise fast and steadily, and we’re pleased to be able to say that sales are really growing, too.
MWJ: Is the Agilent application staff trained in using and supporting both the Genesys and ADS products or do you rely on product experts and a referral system?
TC: One benefit to Eagleware as a result of the acquisition was the access to a worldwide support infrastructure. Agilent has increased the overall level of Genesys support worldwide.
MWJ: I assume this product was developed between the former Eagleware/Elanix team in Atlanta, your EM technology group in Ghent, Belgium that was acquired around 1997 and perhaps even your R&D teams in California. Was the development and integration of this technology among two to three geographically diverse entities a challenge?
TC: Not so much. We’re used to this because we’ve been a multi-site, distributed organization since the early 1990’s (since Hewlett-Packard, now Agilent, acquired the EEsof company). We’ve learned how to do this well, so geography is not a big challenge for us. It also gives us a local perspective and responsiveness that is an asset, not a liability.
MWJ: Todd- I’d like to thank you very much for your time today. The benefits of adding more powerful planar electromagnetic simulation to a product dedicated to smaller organizations seem pretty clear. You have also done an excellent job of explaining how Agilent EEsof has identified the diversity of today’s high frequency electronic design and addressed those needs with multiple products. Many thanks.
My pleasure, David. Thanks for the opportunity!