How Essential Is AutoDesk Products?

[BG2008]

During my 1 year stint as a HVAC engineer, I used AutoCAD to lay out HVAC, fire sprinkler, and plumbing systems that got overlayed with Architectural drawings as well as electrical drawings.  It is the bread and butter of most architectural/engineering firms.  That was almost 15 years ago.  Wow, feels so weird typing it out loud.  I know there are some engineers on this board. 

 

1) Do you use AutoDesk products?  What field? 

2) What products do you use specifically? 

3) Is there a substitute? How are the substitute? 

4) Do you use the subscription cloud service?

5) How important is BIM features and being able to collaborate?

6) What the heck is Rivit? (Not sure if I am spelling it correctly)?

7) Is machine learning designs a big deal now?  In the future? 

8) Are there add ons for AutoDesk that are nice features?  Must haves? 

 

Would love to hear more from this group.  Thanks. 

[chrispy]

I work for a small consulting company doing ship design and would agree - AutoDesk, ANSYS, and similar products are the bread and butter of all firms.  They begin marketing their software to students in university and are absolutely mission critical software. 

 

1) We have a couple of licenses for autocad but use Autodesk TrueView for reading drawings or knock offs such as nanocad or draftsight for the majority of employees to mark up drawings.  One or two employees draft for hours and hours and have the AutoCad licenses.  Autocad is the best of the bunch.

2) Models of designs are continuously sent back and forth with clients.  The client uses AutoDesk software (maybe Revit) to create the model and we view it with AutoDesk Freedom which is free.  This is a very intuitive 3d geometery that can be moved around, walked through, review properties and materials, see if everything fits together, etc.

3) See 1

4) Boss has no interest in cloud due to security and we built our own analysis infrastructure years ago

5) Do not use collaborative software because we can do without and save money

6) See 2

7) I would imagine it is being used in some engineering fields but is non-existent in ship design.  The ship design industry has been around a long time and is slow to embrace new technology and ways of doing things.

 

Note – I purchased shares in ANSYS this morning and believe ANSYS has a more difficult to replicate product.

[Spekulatius]

I am not in Architecture, but the outfits I was working for use Solidworks and sometimes Unigraphics.

[mjs111]

I can weigh in a bit on the Media and Entertainment side, since people have already talked a bit about the Architectural and Engineering side.

 

I've worked in feature film visual effects and more recently at a AAA game company.  These are companies like Disney, ILM, Weta, Sony Imageworks, Blizzard Entertainment, etc.  All of these studios use Autodesk Maya for modeling, rigging, and animation.  They typically have hundreds of licenses per company and have a lot of custom tools built for Maya.  Maya is the industry standard modeling, rigging, and animation package, which means when you hire outside talent that they likely know, and are expected to know, Maya. Conversely, if you have some proprietary animation package, you now would have to train them up on your software, and your software better be like Maya from a UI standpoint if you want that process to go smoothly.  It would be difficult and expensive to switch out your core modeling/rigging/animation package, and there's no great alternative out there with a much better feature set to switch to anyway.

 

If you don't use Maya for animation, you likely may be using its "competitor," 3DS Max.  Autodesk owns that too.

 

Autodesk also owns Mudbox, one of the two industry standard tools for sculpting (the other package is Zbrush). Artists seem to love one or the other here.  Less of a moat on this tool since there are two tools from two different companies of roughly equal feature set out there and it's not a core tool like your modeling, animation, or rendering package.

 

For rendering, Autodesk owns Arnold, one of the industry-standard renderers.  A lot of feature films you've seen have been rendered with Arnold.  Competitors here include Pixar's Renderman, Redshift, and V-Ray, among others.  Big studios will buy hundreds, if not thousands of render node licenses.  The biggest studios, like Weta and Disney/ILM, will write their own renderers, which helps when you scale your render farm out to tens of thousands of processors (Weta's render farm is around 80,000 procs, last I heard). Disney/Pixar/ILM, all of which are owned by Disney, have no explicit license costs when it comes to Renderman licenses, since it's written in-house.

 

Visual effects and game production is a complex project to schedule and keep track of, consisting of many different technical and creative departments, hundreds or even thousands of digital assets and shots, all moving through the pipeline very quickly.  You need something to track all of this. One of the industry standard tools for this is Shotgun. Autodesk owns this too.  Once your studio is set up tracking, bidding and scheduling shows with a particular piece of software, you're not going to switch unless something new comes along that's many times better.  There's just too much infrastructure built up around it to be switching out all the time.

 

 

Mike

 

 

[BG2008]

I can weigh in a bit on the Media and Entertainment side, since people have already talked a bit about the Architectural and Engineering side.

 

I've worked in feature film visual effects and more recently at a AAA game company.  These are companies like Disney, ILM, Weta, Sony Imageworks, Blizzard Entertainment, etc.  All of these studios use Autodesk Maya for modeling, rigging, and animation.  They typically have hundreds of licenses per company and have a lot of custom tools built for Maya.  Maya is the industry standard modeling, rigging, and animation package, which means when you hire outside talent that they likely know, and are expected to know, Maya. Conversely, if you have some proprietary animation package, you now would have to train them up on your software, and your software better be like Maya from a UI standpoint if you want that process to go smoothly.  It would be difficult and expensive to switch out your core modeling/rigging/animation package, and there's no great alternative out there with a much better feature set to switch to anyway.

 

If you don't use Maya for animation, you likely may be using its "competitor," 3DS Max.  Autodesk owns that too.

 

Autodesk also owns Mudbox, one of the two industry standard tools for sculpting (the other package is Zbrush). Artists seem to love one or the other here.  Less of a moat on this tool since there are two tools from two different companies of roughly equal feature set out there and it's not a core tool like your modeling, animation, or rendering package.

 

For rendering, Autodesk owns Arnold, one of the industry-standard renderers.  A lot of feature films you've seen have been rendered with Arnold.  Competitors here include Pixar's Renderman, Redshift, and V-Ray, among others.  Big studios will buy hundreds, if not thousands of render node licenses.  The biggest studios, like Weta and Disney/ILM, will write their own renderers, which helps when you scale your render farm out to tens of thousands of processors (Weta's render farm is around 80,000 procs, last I heard). Disney/Pixar/ILM, all of which are owned by Disney, have no explicit license costs when it comes to Renderman licenses, since it's written in-house.

 

Visual effects and game production is a complex project to schedule and keep track of, consisting of many different technical and creative departments, hundreds or even thousands of digital assets and shots, all moving through the pipeline very quickly.  You need something to track all of this. One of the industry standard tools for this is Shotgun. Autodesk owns this too.  Once your studio is set up tracking, bidding and scheduling shows with a particular piece of software, you're not going to switch unless something new comes along that's many times better.  There's just too much infrastructure built up around it to be switching out all the time.

 

 

Mike

 

Mike,

 

I would like buy you a few beers if we ever meet up.  This is awesome feedback! 

[mjs111]

Ha! 

 

No problem.  Ping me if you have any follow up questions.

 

 

Mike

[coc]

Note – I purchased shares in ANSYS this morning and believe ANSYS has a more difficult to replicate product.

 

What makes you say that? Curious. Thx.

[chrispy]

Mike's response was very interesting, thank you.

 

I should not have said "more difficult to replicate" but should have said "very difficult to replicate" because I am not as familiar with Autodesks full suite. It's very similar to what Mike said but with design and analysis. Say I analyze a structure under 100yr loads. Relevant files and inputs will need to be sent to contractors and government approval agencies such as ABS for ships. These parties will be familiar and have validated the calculations of analysis software like ANSYS. If a lesser known package is used, you'll spend days converting to the proper file formats and defending the validity of the results proving the structure is adequate and safe.

 

ANSYS has a big foothold in electrical engineering which is more of a growth industry then mechanical (unfortune for me). I would read their letter to investors from yesterday afternoon. It gives the sense that they are in a good position despite covid

[BG2008]

Mike's response was very interesting, thank you.

 

I should not have said "more difficult to replicate" but should have said "very difficult to replicate" because I am not as familiar with Autodesks full suite. It's very similar to what Mike said but with design and analysis. Say I analyze a structure under 100yr loads. Relevant files and inputs will need to be sent to contractors and government approval agencies such as ABS for ships. These parties will be familiar and have validated the calculations of analysis software like ANSYS. If a lesser known package is used, you'll spend days converting to the proper file formats and defending the validity of the results proving the structure is adequate and safe.

 

ANSYS has a big foothold in electrical engineering which is more of a growth industry then mechanical (unfortune for me). I would read their letter to investors from yesterday afternoon. It gives the sense that they are in a good position despite covid

 

Just curious why electrical engineering is more of a growth industry than mechanical?  I worked at a HVAC shop back in the days and we had a couple electrical guys in the office.

[lnofeisone]

Mike's response was very interesting, thank you.

 

I should not have said "more difficult to replicate" but should have said "very difficult to replicate" because I am not as familiar with Autodesks full suite. It's very similar to what Mike said but with design and analysis. Say I analyze a structure under 100yr loads. Relevant files and inputs will need to be sent to contractors and government approval agencies such as ABS for ships. These parties will be familiar and have validated the calculations of analysis software like ANSYS. If a lesser known package is used, you'll spend days converting to the proper file formats and defending the validity of the results proving the structure is adequate and safe.

 

ANSYS has a big foothold in electrical engineering which is more of a growth industry then mechanical (unfortune for me). I would read their letter to investors from yesterday afternoon. It gives the sense that they are in a good position despite covid

 

Just curious why electrical engineering is more of a growth industry than mechanical?  I worked at a HVAC shop back in the days and we had a couple electrical guys in the office.

 

 

In the US (and I suspect these trends will hold in most developed countries), there are more mechanical (ME) engineers than electrical engineers (EE) - 55/45 split or so. It's helpful to split out EE field into traditional EE and computer engineers. For MEs and EEs, the majority are employed in engineering services (NAICS 541300) which is basically consulting and companies like Jacobs, Bechtel, KBR, etc.

 

For mechies, the next biggest employment sector is machinery manufacturing, which, in the US, is very mature and is probably in a secular decline. Most mechies are employed in 1) Michigan (auto - declining), 2) California (electronics - growing), 3) Texas (O&G - declining. The growth area for mechies is aerospace but the industry isn't very large and is highly specialized.

 

For traditional EEs the 2nd largest employment industry is Power Generation and Transmission (think utilities). This is one is also mature and stable.  Most EEs are employed in 1) California 2) Texas 3) Michigan. Same industry profile as above.

 

The growth area is really the computer engineering (which bundles everything from systems design to semiconductors). Most of these engineers are employed in California. This group has the largest job growth, highest salary, and higher salary growth (on y-o-y % basis). ANSYS is huge here making the end-to-end process smoother and once embedded, getting an alternative in place requires monumental will and effort.

[clutch]

Autodesk and Ansys are not quite in the same competitive space. Ansys' bread-and-butter is simulation (mainly in mech/elec/manufacturing) and Autodesk is really not strong in that area. Autodesk's main strength is in 3D modeling products for architecture and construction -- namely, Autocad, Revit, etc. They also have a strong presence in media and entertainment via Maya and 3ds Max as another poster pointed out. These products have incredible moats and are not going to be replaced in a foreseeable future.

 

Autodesk is trying hard to get into the mech/elec/manufacturing space with a cloud-based CAD system called Fusion. However, it's mainly used by smaller firms or start-ups. It's primarily a modeling tool but has simulation tools as well -- just not as good as Ansys. Serious and large-scale engineering firms will probably use SolidWorks, CATIA, or NX for modeling and Ansys for simulation. So really, in the mech/elec/manufacturing space the main competitor of Autodesk is Dassault and Siemens. In fact, check out this collaboration initiative between Autodesk and Ansys:

https://www.ansys.com/blog/ansys-autodesk-open-seamless-workflow

 

Another thing to note about Autodesk is that they have been a serial acquirer. Basically, make cash with their cash cows and spend that money to expand. At one point a few years ago, I believe they were second to Google in the silicone valley in terms of the number of acquisitions. They had 100+ smaller products (anything related to 3D design) but they are being consolidated. I think much of the future runaways will depend on how smart they are with acquisitions and in managing their portfolio of products.

 

[clutch]

I hope you had ended up buying this stock...

[BG2008]

Clutch,

 

I did wind up buying some and will likely not sell for a really long time.  Thanks for the feedback on this thread.  After this year's selloff, I used the opportunity to diversify a bit beyond my RE names and bought a bit of tech and other companies that have multi-bagger opportunities. 

[rkbabang]

I'm in IC design and now because of consolidation in the EDA industry just about all of the software we use is made by either Cadence (CDNS) or Synopsys (SNPS).  I haven't used Autocad since I took a drafting class as an elective in my freshman year of college.

 

[Jerry Capital]

Good read on Autodesk here

 

https://www.architectmagazine.com/technology/architects-versus-autodesk_o