The amazing, fantabulous, bewildering world of cloud CAD: application delivery
Invocation of the word “cloud” has now reached a saturation point among CAD circles. Any CAD vendor without some sort of cloud strategy by now would be wise to run -not walk- to the nearest clue dispensing station. Cloud is becoming less of a counterculture alternative, but rather an essential piece of every current and future CAD solution.
But that doesn't mean we're all floating onto one homogenous hipster-laden browser-based CAD nirvana. One size does not fit all in CAD. How you use CAD is relevant in determining how to best utilize the cloud; and there are further tradeoffs to consider.
Cloud ain't just data
If you're just joining the party, last time we talked about cloud CAD infrastructure in broad terms, tackling public, private, cumulonimbus, and everything in-between including the effect of tenancy. But cloud is not simply just about where you might store CAD data, but rather how CAD functionality, or even entire CAD computing environments (complete with compute capacity), can be realized and delivered to users. For CAD, which has traditionally been a heavyweight desktop application, the transition to cloud is a multi-faceted frontier with a variety of potential delivery methods.
We've said it before: not all cloud CAD is the same.
Software update, licensing, and customization dilemmas
The days of software installation media are numbered. Regardless of whether CAD is desktop based or entirely in-browser, software installations and updates can be continuously delivered from the cloud. This one seems like a zombie grade no-brainer, and for the most part it is. So what could possibly be troublesome about having always up-to-date software?
One potential catch specific to CAD: a significant population of CAD users are in fact lazy upgraders. Not because they harbor an unhealthy hatred of software changes (OK there's always that guy) but rather it's about saving the Benjamins when current functionality already meets needs. Which is why so many perpetual licenses of outdated software persist in the wild.
Expect most resistance on this front to be wiped out with aggressive introductory pricing on subscriptions, which for larger companies especially is a more attractive way to expense the costs of engineering tools.
While subscriptions and cloud CAD have been closely coupled from day one, remember that subscriptions are a business model, not a cloud technology. Cloud CAD doesn't require subscriptions, per se. Multi-tenant cloud makes less sense sans subscription, since delivering continuous and global software updates would otherwise be handicapped by complex license management.
Subscription seems to just make the best sense for many (but not necessarily all) business cases both for the customer and the CAD vendor. Like it or not, this is where AutoCAD is going, among others, following in the lucrative steps of Adobe Creative Suite before them. Expect more to follow, and soon.
Subscription or not, another more difficult consideration with cloud CAD is the issue of customization. While a small shop is likely to use CAD as a standalone application drawing files from a network drive, large corporations rely on a variety of customization and integrations with other software. Such situations demand control of how cloud software updates are delivered.
Sometimes this even has regulatory impact, if critical workflows are affected. The chief problem is that providing fine control of how updates are rolled out to individual instances of software kind of undermines some of the efficiency inherent in multi-tenancy. Is there room for single-tenancy cloud CAD or perhaps even variable Windows-10 style policy-driven curation of updates to fill such a niche?
Signs point to yes.
Native apps vs browser
It's easy to think that all cloud CAD must necessarily be browseriffic. And there are very good reasons you might want your next CAD encased in a browser and equally good reasons you might not.
The chief advantage of delivering CAD via web technology like HTML5, is that it's immediately agnostic across operating systems and devices, mobile or otherwise. If the accessibility of your CAD trumps all, than browser based CAD delivery should be near the top of your list of demands. Browser-based CAD can consequently be cheaper to develop because portability of code is maximized which (should) translate to a cheaper product, an advantage easy to rally around. Onshape, for example, has delivered nicely on both of these points.
But there are caveats, in that native apps enjoy a performance advantage over web-driven equivalents. The distinction between natively-coded apps and HTML5 or similar web encodes is less prominent as HTML5 evolves beyond outdated technologies like Flash. For many apps, a web delivery might be perfectly fine, and the user will be happily unawares. However, CAD is rather hefty tool, with hard demands on UX and display technologies. A native client might also provide additional resiliency in environments were online connectivity cannot be guaranteed.
In the content creation industry driven by Adobe creative suite, you'll note their creative cloud is a collection of native apps - chosen no doubt for performance reasons. That doesn't mean you can't take pieces of something like Photoshop, for example, and restructure it into an HTML5 UI presentation for specific uses. In fact Adobe did just that. And just like Photoshop, many of CAD's serious use cases requires native power. But many use cases will be fine in a browser, so once again mileage may vary.
Virtualization and streaming
Just when you thought you might have a good handle on something like native apps, the flexibility of cloud technology throws you another curve... you could have a highly optimized native app performing on beefy hardware in the office, but what if you virtualized that entire environment, put it on some far away space station, and then piped it through via a thin client with remote desktop or streaming technology?
See, this is the kind of thing that makes cloud CAD pretty darn awesome. This approach gives you the same agnosticism on the client side that a browser based paradigm might provide, but with native's ability to not only handle beefy assembly workflows, but also scale further still - to handle visualization and compute tasks simply not possible with conventional (or even slightly eccentric) desktop hardware. That's the kind of thing that is possible with solutions like NVIDIA's GRID, known as a Virtual Computing Appliance (VCA).
VCA tech brings another bonus in enhanced security - since the environment and it's files are virtualized and streamed to the client, the bulk of the IP (save the display) stays at the remote location, so wholesale export of valuable data is essentially impossible. That's a rather handy security bonus.
But like with everything else cloud CAD, there are limitations here as well. For one, such implementations are rather expensive for the time being especially if you include the cost of having a sufficiently fat internet pipe. There are also two prime adversaries than can create a less than desirable end user experience: streaming bandwidth and latency. You're not going to be doing this from a Starbucks anytime soon.
But perhaps some day when you're nursing a sippy cup at a distant future Taco Bell, internet will be pervasive enough. Maybe.
The central theme point to take away here is that cloud CAD is more about choices than absolutes. Where decisions about how you use CAD matter are central to how cloud will factor in your future cloud software. Balancing cost, user experience, portability, performance, security, are all highly relevant questions for CAD. Which way is best? You may have your own opinions. Time to express them.
We're making professional 3D printing easy and accessible
There's a new solution coming your way this summer. Click here or on the graphic to the left to request updates and beta access.
About the author: (Ed Lopategui)
Technology evangelist, entrepreneur, and aerospace engineer specialized in the software tools and technology which enable engineering, design, and product development - PDM, PLM, CAD, CAE, CAM. Any views, opinions, prophecies, and sarcastic remarks are my own and are in no way associated with any current or past employers.
All posts by Ed Lopategui — Follow on Twitter