I always believed that I was interested in electronics as a kid. I now see that I was wrong. After chatting with Ken McElvain at this year’s DAC, I learned that I was merely a poser, dabbling in the discovery of technology. Ken McElvain was interested in electronics. While I sat in my room with my Radio Shack 100-in-1 kit hooking up the multi-colored wires to the spring-loaded terminals, carefully constructing every pre-made project in the book, Ken McElvain was in the backyard with his dad, blowtorching TTL components off discarded circuit boards to stuff into wire-wrap sockets, making his own early digital designs. While I was building primitive photo-cell circuits to alert me when my little brother wandered down the hall toward my room, Ken McElvain was in the fourth grade building his own adder out of electromagnetic relays. When I heard this story, I was puzzled. My Radio Shack kit had only one relay – not enough to build an adder. “I wound my own.” Ken said.
Almost three decades later, when Ken and his wife Alisa Yaffa founded Synplicity, he was still “winding his own.” Synplicity started from modest beginnings with Ken and Alisa working from home. Alisa developed the business as Ken worked solo writing the first version of Synplify, the most widely used commercial logic synthesis software in the world. Bucking the predominant EDA trend of acquiring venture capital based on PowerPoint slides, then trying to develop technology later, Ken and Alisa self-funded the company for two years, and had the first customers successfully synthesizing away before they went for additional expansion funds.
Growing up, Ken’s passion propelled him to take college-level and later graduate-level computer science courses at Washington State University in the afternoons while attending high school in the mornings. During this time he developed his first logic synthesis tool – a state assignment program. He later attended Washington State as a full-time student, and then went to work designing processors at Hewlett Packard in 1980. Ken lived a dual life during this time, designing an ECL mainframe and developing EDA tools to support the design. The computer family was later canceled, but the tools were used in the design of the first HP Precision Architecture computer, which eventually led Ken into the EDA industry.
While Ken’s reputation as a technologist is somewhere between well-known and legendary in the EDA community, the technologist title sells him short. Ken is an innovator. Ken’s innovation is obvious in the technology he developed and shepherded, but it is also apparent in the approach he took to creating an EDA company that breaks the mold, structuring his business to succeed in a new and expanding market.
At the time Synplicity began, most FPGA and CPLD designs were still done with schematics. The leap from schematic-based design to hardware description languages and synthesis was a significant barrier for most programmable logic designers. Ken set out to make that transition as easy as possible by producing a product that bucked the trend of feature-rich synthesis tools popular in ASIC design. Synplify was designed to have the absolute minimal set of controls and options for a designer to learn. “I still talk today about the true cost of adding a feature,” says Ken. “Every control or option you add represents another possible wrong path the customer can take, and an additional training and support burden. We want to produce products that work well right out of the box.”
Synplify brought another trick to the party as well. It was fast. It did the synthesis job in far less time than any tool on the market, reducing the critical iteration time in the debug-synthesize-layout loop. “For our customer, faster tools meant fewer tools to get the job done,” says McElvain. Synplicity built a value-based argument for using their tools over those supplied in the FPGA vendors’ design kits. Performance, quality of results, stability, robustness, and vendor independence all worked to justify the cost of a commercial EDA tool like Synplify over the “default” vendor-supplied one. In some cases, Synplify even became the vendor-supplied tool through OEM agreements. “We have always realized that we have to stay ahead of what the vendor tools can do,” McElvain continues. “If people could get the same capabilities with a cheaper tool, they’d do it. We maintain our position by continuously improving.”
Synplify was wildly successful in the FPGA market with its pushbutton ease-of-use, its value-based pricing, and its responsive sales and support model. As design teams converted from schematic-based design to HDL-based design, Synplify became the synthesis tool of choice for designers making the switch and wanting simple, high-quality, vendor-independent synthesis.
Ken knew that the FPGA market could not be won with a traditional style EDA sales channel. Normally, EDA is optimized to sell low volumes of high-priced, high-support products to a small number of large customers. For FPGA, the model needed to be almost the opposite. The sales channel had to be nimble, aggressive at lead development, and able to handle large volumes of smaller transactions. The product also had to be easy to use and support as the large number of installations could quickly overwhelm a traditional EDA support organization if the tool were high-maintenance.
Synplicity also recognized the value of partnering with FPGA vendors from the beginning. They quickly established close ties; both on a technical level, and in marketing and sales to leverage the mutual benefit of Synplicity’s tool technology and the FPGA vendors’ silicon and access to the customer base. A lot of thought went into building the right combat tool for the FPGA vendor’s field applications engineers (FAEs), starting with a user interface that fit on the 640×480 screen available on early laptops. If the vendor FAEs were able to use the tool to make their customers successful those customers would probably want a copy for themselves.
It was obvious from the beginning that Ken planned to skip the usual EDA startup tradition of technology development followed by acquisition. By developing an innovative product for a new market, an innovative sales channel, and new business relationships outside of the usual EDA realm, Synplicity captured a dominant share of the emerging FPGA tools market, and went public in October of 2000, just as the downturn was getting started.
Having established a track record in FPGA synthesis, they also began to expand into additional adjoining markets as well as creating new tools for their existing customer base. Realizing that many ASIC teams were starting to use FPGAs as prototyping vehicles, they created Certify, which helps convert and partition ASIC designs into FPGA prototypes, reducing the risk and effort of verifying large ASIC designs. They also realized that physical effects were becoming a problem in FPGA as densities grew and geometries shrank, and preemptively launched their Amplify physical synthesis product to help high-end FPGA designers cope with the challenges inherent to larger, more complex designs.
With the FPGA tools market showing a steady but bounded growth, Ken was looking for new, larger markets to diversify the company’s customer base while leveraging its technology and talent resources. The trouble was that the logical step was into ASIC synthesis, a market already dominated by a large entrenched supplier. More interesting and better matched to Synplicity’s strengths was the newly emerging structured ASIC market. Designers looking for a happy middle ground between the fast, low-cost design cycles of FPGAs and the high-density, high-performance characteristics of ASIC were looking to the new structured offerings. Synplicity saw an opportunity to jump into a leadership position in the new market, serve the larger, more diverse customer base with their well-adapted sales channel, and catch the larger, less nimble EDA vendors sleeping when structured ASICs began to hit their stride.
Most recently, Synplicity went after the rapidly growing DSP-on-FPGA segment with their newly announced Synplify-DSP tool that provides an entry path from MATLAB/Simulink into the traditional FPGA design flow. This represents yet another potential expansion of the customer base while leveraging the core technologies and strengths that have brought the company to success thus far.
Today, with cell-based ASIC design starts declining, and FPGA and structured ASIC starts on the rise, Synplicity is well poised to capture an increasingly large share of the EDA market. As the larger EDA vendors start to respond to changing market conditions by introducing new tools aimed at the FPGA and structured ASIC markets, they are finding that Synplicity is already there, and in many ways is better adapted with a more nimble sales channel and more supportable and approachable product offerings.
Being ahead of the game isn’t new to Ken McElvain, though. Throughout his life he’s been at least a little ahead of the crowd in moving to the next thing. He could’ve told you that back in the seventh grade, in the backyard, extracting TTL parts from scrap electronics boards with a blowtorch…