Synopsys with Ansys may be too big to be acquired by a design software company.
Synopsys’ acquisition of Ansys for $35B, expected to close the first half of 2025, will be the biggest deal ever in the design and engineering software industry, dwarfing the $4.5B acquisition of Mentor Graphics by Siemens in 2017.
Until now, simulation software has been increasingly integrated with design software. Ansys had declared itself a technology partner of Autodesk despite the fact that Autodesk had already purchased a bevy of simulation software companies — including Blue Ridge Numerics in 2011 for its computational fluid dynamics (CFD) software and NEi Nastran in 2014 for its finite element analysis (FEA) software. Ansys CEO Ajei Gopal visited Autodesk CEO Andrew Anagnost at Autodesk’s San Francisco headquarters in May 2023. They also shared the stage at Autodesk University in 2019. It’s enough to make one think that there was something more binding in the offing.
But Ansys was playing the field. A Siemens/Ansys partnership was announced in February 2021 despite Siemens having a division devoted to engineering design and simulation. PTC proudly announced that it would be using Ansys for simulation inside Creo.
In fact, many simulation-software companies have been acquired and absorbed by design-software companies. All but the largest were acquired and had their technologies “dumbed down” and served up to an audience of mostly mechanical designers and engineers. Mechanical engineer and SolidWorks founder Jon Hirschtick realized how critical it was to integrate analysis with mechanical design, so acquired SRAC for its COSMOS FEA program in 2014. Then CD-adapco was acquired by Siemens. Autodesk, as mentioned above, picked off several second-tier simulation companies. One of the largest and best-known FEA companies — MSC Software, makers of Nastran — was acquired by Hexagon.
By the end of 2023, the only independent simulation-software companies left standing were Altair, COMSOL and Ansys. While Altair and COMSOL could be considered too aloof or too independent, Ansys was just too big to be acquired. Hexagon paid $834M for MSC Software, but CAD-company deals for simulation companies were for orders of magnitude less in value. Ansys had a valuation in the billions of dollars. No CAD company could buy them, right?
The trend to integrate simulation
With each acquisition and partnership between design and simulation software companies, the trend to integrate simulation into design became more apparent. The prevailing wisdom with design software companies was that simulation ought to happen early in the design phase rather than after it. Simulation was being made pushbutton easy and put in the hands of designers, some of whom couldn’t spell FEA or CFD, much less understand its underlying theory. “Democratization,” coined for bringing CAD to the masses, reared its ugly head again for simulation. There was much said about putting simulation “under the hood,” never to be seen in setup or execution, only for results.
If Ansys by itself is too big to be swallowed by a CAD company, there is little hope of the combined Synopsys/Ansys being acquired. Combining giants of MCAE (mechanical computer aided engineering) and EDA (electronic design automation) will make a titan too big to be acquired by any existing design-centric software vendor.
It would’ve been more aligned with the way analysis was being blended into simulation if Ansys, the MCAE leader, would’ve merged with a leader in design software. Then the melding of design and simulation that needed to happen could have progressed.
Another deal that would have made sense would’ve been the merging of a leading CAD company with an EDA leader akin to the Siemens acquisition of Mentor. After all, mechanical design must be integrated with electrical design.
Instead, we have an isolated titanic software company created from a mashup of two companies no one asked for, not even their users, a company too big to be acquired by other companies in this world of engineering and design software. Synopsys may end up the big kid on the playground no one plays with.
A big deal but is it a good one?
The Wall Street Journal broke the story in December, but the companies took until January 18 of this year to go public.
Ansys 2024 revenue was $2.16B in the 12 months before September 30, 2023. The $35B Synopsys is offering for Ansys is a whopping 16 times Ansys’ annual revenue.
Ansys was at one time the fastest growing company in the design and engineering software space, the envy of other design and other simulation vendors. Ansys routinely delivered above-market growth, growing from $50M annual revenue to a billion dollars in 16 years. Its share price climbed to nearly $400. Since such lofty heights, the company has suffered from wild fluctuations in share price and lackluster performance over the last several quarters — factors that may have contributed to make Ansys ripe for acquisition.
Meanwhile, Synopsys has been on a tear. SNPS shares on the NYSE are up nearly 55% to about $542 apiece over the past year.
In the fourth quarter ending Oct. 31, 2023, revenue was $1.6B, an increase of 25% from the same period the year before, and in the 12 months ending Oct. 31 the company brought in record revenue of $5.8B, an increase of 15%.
Who wins with this acquisition?
Shareholders of Ansys stock are clear winners, as Synopsis is paying a premium for their shares.
Synopsys stands to broaden its customer base with the Ansys acquisition. Synopsys, a purported leader in semiconductor EDA, has no doubt ingratiated itself with big, established chip manufacturers — the few that there are. With the Ansys acquisition, Synopsys is in a more favorable position for courting big manufacturing companies such as automakers and aviation, transportation, and heavy machinery manufacturers. In that world, Ansys is well known and respected. And in this world there are increasingly computerized designs reliant on overseas semiconductor suppliers.
Manufacturers had quite a scare when supply chain issues stalled assembly lines during COVID. The fear continued with the mounting tension between China and Taiwan as Taiwan is where most of the world’s most advanced semiconductors are produced. Synopsys offers them a solution: Why not make your own semiconductors?
Chip design and manufacturing is no longer the sole premise of giant chip foundries, such as Intel and Taiwan’s TSMC. Chip customers are looking to bring chip design in house, taking a cue from Apple. Apple took on making their own microprocessors with the M series SoC and now have the fastest and coolest (as in low temperature) microprocessors for their computers and devices. The increase in cloud storage, AI, computerized vehicles , and smart products have all contributed to the demand for chips. The anxiety over the supply and the dependence on foreign sources have made national governments quite concerned and ready to fund their domestic production. The stage is set for in-house chip design — and for chip design software vendors to cash in.
Synopsys’ good fortune is shared with its chief competitor, Cadence Design Systems, which has also experienced heady revenue increases in the last year. The two companies combine for the majority of EDA for chip design done worldwide.
MCAD and EDA siloes
Mechanical CAD (MCAD) and electrical design automation (EDA) software developed separately and have continued to operate separately, each in their own worlds and each with their own factions. Despite products becoming increasingly electrical and computerized, MCAD and EDA are tools of distinct disciplines, respectively, serving mechanical engineers and electrical engineers. There have been attempts to bridge the two, such as with software modules in MCAD platforms, and notably Autodesk’s Fusion 360, which includes EDA. Presumably, EDA platforms have apps that do rudimentary mechanical design and thermal analysis.
With both EE and ME curriculums stretched to the point of breaking, there’s little room to add courses from one into the other. My BSME program included chapters on electricity and magnetism in a physics series and one course in circuits, which with its imaginary numbers might as well have been taught in a foreign language. A predilection for all things mechanical was to relegate me to second-class citizenship in my first engineering position with a defense electronics company. We were a small team of mechanical engineers among hundreds of EEs who got the glory for millimeter wave antennas, electronic countermeasures and advanced communication systems. We got to package, test, and ship it all.
Getting the heat out
But with the increasing miniaturization of circuits, the stacking of dies results in a thermal density so high that the lack of heat dissipation becomes a design-limiting factor. To devise ways to manage the heat, EEs and MEs shuttle the design back and forth: EEs do the chip design, then MEs do a thermal analysis and send it back to the EEs for a redesign to correct troublesome hot spots… an iterative process.
A computer workstation’s microprocessor is easily the hottest thing in a desktop chassis, and in a cramped mobile workstation, it gets hotter yet. Intricate cooling systems with fans and air vents, copper conduits filled with fluid and even immersion of the microprocessor, are all methods used to prevent overheating.
Even if the hot spot temperature is lowered to allow devices to continue functioning, the heat produced by electronics is a problem at board, system, rack, facility and environment levels. Excessive heat produced by servers is a challenge for facility designers who have to create special liquid cooling and specify enormous air conditioning systems. Then if you add up heat given off by all the data centers built or being built because our fondness for cloud computing, you have a factor in global warming.
The end of rinse and repeat?
At present, the electrical engineer must create chips, boards and systems and send the designs over the wall for a thermal simulation, a process that may be repeated several times. In the future, an integration of MCAD software and ECAD (or MCAE and ECAD such as ANSYS and Synopsys) would yield benefits in one of two ways.
It would shorten design cycles by reducing if not eliminating the number of iterations between EEs and MEs with the EE handling it all.
It would allow for design optimization with more iterations. In the same time it took separate EDA and MCAE systems to independently iterate, and integrated EDA/MCAE platform could do many times the iterations for a survivable design that’s also optimized.
You may also like:
Filed Under: Commentaries • insights • Technical thinking, ENGINEERING SOFTWARE