Why CAD is not moving to the cloud (quite yet)
My last post was about the integrations of different SaaS applications. My pen-friend Oleg shot back with another article – Why cloud won’t solve CAD/PDM/PLM integration problem tomorrow? He made a very interesting point about integration – that the last product moving to the cloud is the CAD software. Is this true?
I want to take a closer look at the reasons engineers typically give for wanting to move to the cloud.
I speak to at least a dozen engineers a week and one thing I keep hearing is that they love their CAD tool. That’s not to say they like the zero-value subscription business that comes with it (as discussed on SolidSmack), but they love designing with their CAD tool. They have spent years to master their software and they can’t live without it. Given this, what would cause CAD to move to the cloud and what would hold it back?
Arguments for moving CAD to the cloud
1) I don’t like paying subscription fees.
CAD users have to pay ridiculous subscription fees to get service packs. Yes, you need to pay for a software vendor to actually fix the bugs. But the other option is spending thousands of man-years moving CAD to the cloud.
Autodesk and Siemens have introduced subscription pricing for their CAD products. Although the price for these tools may not be attractive today I suspect they will be lower in the future. All it takes is for one of them to price aggressively and the other vendors will follow.
2) I can’t share my files with anyone outside my company- like suppliers.
It’s difficult to share CAD files with suppliers and to access the files anywhere. Obviously I’m biased (see GrabCAD Workbench to understand why), but I think this is a big cloud opportunity for CAD and managing collaboration.
Compare CAD to solutions like Evernote or Adobe. Even though Evernote and Adobe are installed tools, you can sync files to the cloud and access them from other locations. CAD itself doesn’t need to be on the cloud to achieve the same thing.
3) My team needs to make edits simultaneously.
On the cloud, contributors can edit the same file at the same time just like Google Docs. CAD should be on the cloud so everyone can make edits to a file whenever they want.
If you look at how CAD users work, it’s actually more like software engineers, not lawyers. Engineers make edits to parts and subassemblies then merge them to the top level assembly. There is a process around it. Someone needs to check to make sure that all the changes make sense. You don’t need to put CAD on the cloud to make that work – instead you can build a tool that helps to manage changes. As an example, software engineers use on-premise tools for authoring and cloud tools like GitHub to manage their work.
4) I want infinite compute power.
There are certainly benefits to “elastic computing” – expanding the number of machines you are running on. CAD should also be able to take advantage of parallel computing on the cloud.
Unfortunately, the cloud makes CAD slower, not faster. Solid modeling kernels require very fast single processors, so your best bet for speed is to buy the fastest processor you can. There are also difficult issues with latency that can make the interactive experience inferior to desktop tools. That said, other engineering tools may benefit from elastic computing. For example simulation and rendering have many compute-intensive operations (that’s why big companies have supercomputers running on their farms). These are perfect candidates for the cloud and we will probably see many of these players moving there.
So who goes first?
As you can see, I think that for many traditional CAD users there aren’t overwhelming reasons to look for a cloud-based CAD solution. But CAD users aren’t heterogeneous – there are many users for whom a cloud-based CAD solution makes sense.
- People who are just getting started – If you don’t have a favorite CAD system already why not try one that doesn’t require upfront fees and ongoing upgrades?
- Users who don’t need all the bells and whistles – Desktop CAD systems are amazing software, with amazing features, but not everyone needs that much power (and the complexity that comes with it).
- Folks who want to edit but not create – Some users just want to tweak a model, not start from scratch. A more accessible cloud-based CAD system may be a good fit for them.
I agree with Oleg that for most users CAD tools will be among the last to move to the cloud. Many of the benefits of cloud computing are limited for CAD (as it stands today). But the opportunities for the tools “around” CAD are quite significant. I think those tools will move to the cloud first, and CAD will eventually follow.
PS I wrote this post on the plane using evernote (installed on my laptop), not on my cloud-based google docs app.