If ever there was a free thinker in this industry who isn't afraid to drive the bus in a new direction, it is AWR's CTO Joe Pekarek. Joe founded this company on personal credit cards because he was convinced there had to be a better way to design than what the industry giants were offering. Innovation is a key element in the "drive the bus" leadership mentality, and AWR fully embodies that concept. From its first innovation, the unified data model (which did, in fact, offer a better way to design), to its latest technological breakthroughs -ACE™ and AXIEM™- AWR has broken new ground in offering unique design tools that help customers do their jobs faster and better.

Driving the bus encompasses more than delivering new technologies - it also includes innovative and forward-thinking partnerships and initiatives. Whether it is AWR's founding member role within the Interoperable Pcell Library initiative (IPL) or our recently announced Project Jumpstart with TriQuint, AWR is leading customers and the industry down new paths with the reward of increased productivity.

Lastly, every now and then we should not hesitate to stop driving the bus to refuel and recharge. So mark your calendars now for AWR's 4th Annual Customer Appreciation Party at the Hard Rock Café in Altanta during the week of IEEE-IMS (MTT 2008).

You joined AWR in 2003 - what differences do you see in the RF & Microwave EDA industry now compared with five years ago?

The biggest difference is the consolidation within the industry. In 2003 there were about six players in the circuit simulation market but in the last three years, AWR has bought APLAC, Agilent has bought Eagleware and Xpedion and Ansoft has focused on the 3D EM market. The current situation now is that over 95% of the core frequency domain circuit simulation market is shared by just two companies.

What has been the driving force behind this consolidation in your opinion?

Technically EDA acquisitions tend to be made to get closer to the Holy Grail of a "complete solution" where one product meets all requirements for all customers. In reality this will never happen but there is no doubt that most users would prefer a single vendor solution if one existed and so EDA vendors strive to achieve it.

For example, a major reason for AWR's purchase of APLAC was that although AWR had a harmonic balance engine that worked well for most non-linear problems, the APLAC harmonic balance engine had greater capacity and was the fastest on the market, especially for larger circuits. Harmonic Balance was viewed as a core technology by AWR and fundamental to the company's growth so the decision was taken to buy APLAC, integrate the APLAC harmonic balance simulator into the AWR GUI and relocate development of non-linear circuit simulation to the AWR-APLAC site in Finland.

AWR just finished its Financial Year 07/08. Was it a successful year for AWR in Europe in your opinion?

I think it must qualify as a successful year in anyone's opinion! AWR bookings in Europe were up by over 30% on the previous year and are now at a sales volume more than 4x higher than when I joined AWR in 2003. We had record bookings in our four largest territories - the UK, France, Germany and Scandinavia - so the success was shared throughout Europe. Globally bookings were up by more than 20% despite tough economic conditions in some regions, especially the USA.

You described the EDA market earlier as "fairly static". Why do you think AWR continues to grow so rapidly?

This is actually our sixth consecutive year of double digit growth and the main reasons are the same as for most successful companies: good products, good support, continued innovation and competitive pricing.

On the product side we have significant technical advantages over our competitors in terms of speed, accuracy and capacity but generally the feedback from our customer is that our biggest single advantage is the intuitiveness of our product --especially for the infrequent user.

Support is another critical area in EDA software and one in which we have invested heavily. We have local Applications Engineers in every major territory to support the users and additionally a dedicated post-sales support team to deal with more time-consuming, consultative issues.

Customers like to see innovations that help them do their job and in the last year we have introduced both ACE, our RF circuit extraction tool and AXIEM, our next generation 3D planar EM simulator. ACE is a unique product, taking extraction techniques previously only available at very low frequencies and adapting them for use in the RF and microwave design arena. This technique offers the accuracy previously only available with EM simulators but in a simulation time orders of magnitude less.

What do you expect to happen in the industry over the next few years?

There is little doubt that the economy in general is set for a bumpy ride over the next couple of years. Tighter controls on consumer credit will mean a reduced demand in many market sectors. In general this leads to a reduction in capital spending by many companies. My experience is that under these circumstances there is even more pressure than usual on engineers to justify any purchases they want to make i.e. do we really need this? So as well as technical benefits I expect there to be even more focus on the improvement in design efficiency and the tangible time and money savings that result from using EDA tools.

Aside from the economic outlook, the focus of the EDA industry for the last few years has been to take existing "point tools" already developed and link them together in some way so as to form a complete design flow. This has been one driver in the recent flurry of acquisitions.

AWR addresses this issue by providing an open "socket" to allow other vendors' products to be used seamlessly with its own products and it is likely that this will be the future direction of the EDA industry. The ultimate solution for the customer is the Open Access framework which allows the user to link any EDA tools from any vendor into the design flow.

This is not an uncommon problem. The designer can often feel as if caught between the proverbial rock and a hard place. On the one hand, EM analysis, when used properly, can give a precise model for the interactions but provides little insight as to which bits of metal are causing the problem. On the other hand, the time and chances of making a mistake are prohibitive if you were to put together a detailed schematic using coupled lines models, junctions, bends, etc. So what's an engineer to do?

One alternative is to use AWR's ACE™ Automated Circuit Extractor technology as both a gross and fine check of interactions. By way of an inverse example, we will use Distributed_Amplifier_ParLayout_Extract.emp to look at how the interactions among the trombone sections (shown in Figure 1) contribute to the overall gain bandwidth. Figure 2 shows the baseline performance taking into account all couplings (ACE coupling radius set to 500um).

As a gross check, we can slowly change the ACE coupling length to alter the complexity of the coupled line structures in the extractions. By reducing the ACE coupling distance to 150um (illustrated in Figure 3), we can see the dominant couplings that contribute to the full bandwidth performance of the amplifiers.

For a fine check, we may wish to do some optimization or sensitivity analysis using ACE at the 150um coupling distance to see how critical these couplings are. One way to go would be to use manual tuning to change the separation between the first, highly coupled trombone section to the input lines. Another way to go would be to put in process variants for line width/spacing and see if this causes a large variation. Other alternatives are also possible given that you can look at the exact netlist and model parameters that ACE has extracted.

As for the problem with your co-workers... I suggest you bring in donuts once or twice!