Semiconductor EDA Face-off: Cadence vs. Synopsys

In a screen-based investment strategy like we use here, it isn't unusual to see many competing firms show up at the same time.

Despite a lot of time and effort put into researching individual companies, the truth is that underlying industry dynamics account for a lot of any particular company's success or failure. This is particularly true for the "established players".

Recently, I've noticed two very similar companies enter the Green Screens: Cadence Design Systems (CDNS) and Synopsys (SNPS).

This is intriguing. These two companies dominate the semiconductor electronic design automation (EDA) software space, each with about 30% market share. The only other firm even close is Siemens EDA, formerly Mentor Graphics, at about 13% share.

In this article, we're going to take a look at the underlying industry itself to gauge its investment potential, then we will pit Cadence and Synopsys in a one-on-one "death match" to see which is the better Watch List candidate - if at all!

What They Do

The similarities between these two firms are pretty remarkable. They are the "Coke" and "Pepsi" of the EDA industry.

The bulk of both firm's business - 55% for Synopsys and over 85% for Cadence - is selling software used to design and validate integrated circuit (IC) products. Every electronic chip in every single device ranging from laptops to cellphones to smart speakers and automobile control systems (just to name a few) goes through the same process. It first must be designed and validated to work from a logical standpoint, then that logical design must be converted to a physical layout, which also must be validated against manufacturing, electrical, material, and other rules. All the while, concerns about ease-of-manufacture, size of the chip, operations in target temperature and voltage scenarios, and so forth must be considered.

Synopsys and Cadence both provide software tools that help IC design houses accomplish these goals. Virtually any company designing chips use one or both of these vendors to get their designs through production. These are absolutely critical tools. The earlier design defects are caught, the less expensive they are to fix. The absolute last thing an IC maker wants are defects getting through to wafer production, where they become EXTREMELY expensive to address.

The second piece of business for both firms is selling pre-created blocks of IP that can be "dragged-and-dropped" into a customer's chip design. For example, a customer may be designing a product that needs to communicate via a voice operated user interface. In this case, instead of designing their own voice recognition and synthesis chips, the customer could simply buy a pre-created analog-to-digital (ADC) block from Synopsys or Cadence and drop it in to their design.

For EDA software, both companies use a primarily "time-based license" model. Customers enter into license agreements for a certain set of tools over a 2 or 3 year period. One important thing to note is that these are not your typical "software-as-a-service" subscription models. The data used in these tools is often far too big to be managed in a cloud-deployed model, although some functions are moving to the cloud for both firms.

On the IP side, revenue is collected in one of two models. Sometimes IP is licensed by a customer for indefinite use, and sometimes there is a royalty-based model where Cadence/Synopsys captures a certain dollar amount per part shipped.

Revenue Outlook

As these two firms dominate the space, the growth outlook for the industry basically amounts to the growth outlook for both of these companies.

From a numbers perspective, semiconductor EDA was roughly an $11.5 billion dollar market in 2020. It is expected to grow to $18.5 billion by 2026, representing an 8% annual growth rate. I see both of these firms out-pacing that figure by at least a few percentage points, given their effective duopoly and "trusted partner" status.

Synopsys has grown sales a bit more rapidly than Cadence, with a 15% 3 year CAGR, vs. Cadence's 12%. On the other hand, Synopsys has also been a far more acquisitive company in its history, sporting almost $3.9 billion in goodwill assets vs. $1.3 billion for Cadence. As a result, Cadence has a more attractive cash return on investment figure, at 39% vs. 25%. So, in a nutshell, Synopsys has grown a bit faster via acquisition, but Cadence has been more efficient in its growth strategy.

Either way, it's pretty clear that computer chips are going to continue to be a key component of nearly all products going forward. Chip shortages stemming from pandemic supply chain issues were a massive problem in 2021, as chips are now a key part of everything from automobiles to refrigerators! I don't see any major concerns in either of these companies hitting 10-15% revenue growth targets in the near-term.

The last thing we look for is recurring revenue. Both of these companies tout 85-90% of their revenue sources as recurring. EDA software licensing is a pretty steady business, as new integrated circuit products need to go through the same design and validation processes over and again.

Is There An Economic Moat?

The next question to ask is: do these firms have durable competitive advantages, or an economic moat, that protects them against new competitors?

In fact, I see several moats at play here.

The first are very HIGH SWITCHING COSTS. Building a chip design and validation flow from logic synthesis all the way through to physical verification is highly complex. Engineers from several different parts of the product chain need to collaborate and share information. Both Synopsys and Cadence have developed full end-to-end flows that make this easy for its customers. IC design firms tend to standardize on one or the other, and switching would be HUGELY disruptive!

The second moat factor is customer trust, which falls under the BRAND advantage. Catching logical and physical flaws before a chip goes to fabrication is absolutely critical, and the earlier defects are caught, the less expensive they are to fix. If you are a product manager, are you going to trust Synopsys and Cadence or some new, cheaper EDA vendor trying to get established in the market? It's an easy choice. Paying a little more for a trusted name could save you big bucks in the long run.

Finally, we come back to the end-to-end, full flow software platform offerings both vendors provide, which amount to a UNIQUE ASSET against most of its competitors. Many small EDA start-ups offer "point tools" that focus on one or a few parts of the design and validation flow, but almost none can offer end-to-end integrated flows. Combined with the relatively small size of the EDA market at large, this creates a "big fish in a small pond" scenario that discourages new competitors from even entering.

If you look at the history of semiconductor EDA going back to the late 1990's, you'll see the same 3 names as leaders: Synopsys, Cadence, and (kind of) Mentor. That's right... despite all the technological advances over those 20+ years, these 3 firms (really 2) continue to lead the pack! The above moat factors are why. I see no reason to think it won't continue.

Leadership and Financials

The management story is a bit different for these two firms.

Synopsys is still led by co-founder Aart de Geus, who has served in a CEO capacity since way back in 1994, and chairman since 1998. Throughout his prestigious career guiding Synopsys, he has won several industry awards including CEO of the year in 2002.

Cadence, on the other hand, recently went through a CEO transition, with Anirudh Devgan taking over for Lip-Bu Tan, who had held the chair for 14 years. Devgan has been at Cadence in high-level executive positions since 2012, and before that at Magma Design Automation (which was eventually purchased by... Synopsys!).

While readers know I always lean towards co-founders, I actually don't see much of an advantage on either side here. While de Geus is a known quantity, at 68 years old he is likely nearing the end of his leadership tenure. I was also surprised at how small de Geus's ownership stake in Synopsys is. Despite being a co-founder and with the company for nearly 40 years, he owns well less than 1% of the stock.

Financially, both companies are strong, as we've touched on earlier. Their free cash flow margins are almost identical at an excellent 30%. Cadence's ROIs are slightly better due to a more organic growth strategy. Synopsys is basically debt-free, and Cadence maintains a very manageable debt load that is under 30% of equity (very safe territory).

Put simply, neither company is clearly better than the other when it comes to leadership or financial strength.

Risks

It is pretty remarkable that these two firms have managed to stay on top of the heap given the rapid advances in semiconductor technology. As chips have gotten ever smaller, problems as basic as the physical width of a beam of light have become challenges in manufacturing!

This march of technology generally gives new competitors ample opportunities to unseat established players. While we haven't seen that happen in EDA to date, it is still a big-picture risk for both firms.

There have also been some regulatory concerns. Last August, both companies were ordered to stop providing services to Chinese firms developing AI and high performance computing chips. The recent spy balloon debacle isn't going to help U.S. wariness of Chinese ambitions. Still, sales to China are under 15% of total revenues for both Cadence and Synopsys, and the bulk of those don't involve high-end designs. Additionally, both firms would benefit from a increased build-out of chip infrastructure in the U.S.

Finally, the volatility of the semiconductor space historically has been a concern. This has also eased quite a bit over the last 10 years. Prior to that, the industry was driven mainly by PC and electronic sales, which are highly sensitive to economic conditions. Now that chips are found in far more use cases, this volatility should be muted somewhat. Both companies would be hurt in a recession, but not substantially more than the average firm.

Which One To Buy?

We've established a pretty good case for both Cadence and Synopsys. They are nearly a duopoly in a critical, growing industry. They have strong moats in switching costs, branding, and hard-to-match unique assets of full end-to-end EDA flows. Both companies report 85%+ recurring revenue.

Frankly, I think either company is worth slotting into the Watch List. Cadence is more attractive from an ROI standpoint. Synopsys has grown a bit faster in the past and has more seasoned management. Really, this is a toss-up from a business perspective.

The winner here will be Synopsys, and it will be added to the Watch List today. There is one simple reason why... Synopsys' stock looks much more reasonably valued than Cadence at present (neither look especially cheap). We will keep an eye out for an attractive buying price in the future.