Seems nowadays its hard to read about engineering IT without coming across something about the cloud. And while it’s definitely an interesting concept, it hasn’t necessarily been explicitly clear what advantage or value it providers to designers, engineers or analysts. With ideas of IT related issues that have plagued engineering in the past, I started to think hard about the cloud.
Background on the Cloud
I won’t go into a great detail about what the cloud is here. That wouldn’t be adding much value to the discussion. Instead I’ll point you to what I rely on in terms of understanding the concepts. Technewsdaily posted an excellent graphic showing that Your Computer is Going Away. This one’s easy to skim and grasp the concepts. For a text-based definition that goes into greater detail, here’s the cloud entry on wikipedia.
The Advantages and Disadvantages of the Cloud
With some definitions out of the way, let’s talk about the good and bad of the cloud. As I see it, there’s three general capabilities that warrant some discussion.
- Processing: One of the more enticing aspects of the cloud is the idea that you can access more powerful processing capabilities, whether that be general processing, graphical processing or RAM, quickly and easily. The advantage is to executing computationally intensive activity far more quickly.
- Storage: Another attractive characteristic is the ability to expand your storage capacity as necessary. It’s something to considering given the rate at which manufacturer’s data stores are expanding.
- Data Transferral: One potential downside of the cloud is the fact that every activity is dependent on network bandwidth.
But that’s not the whole story, is it? The cloud for enterprise systems (meaning servers) offers some advantages that are different compared to the cloud for software applications (meaning desktops). Let’s talk about that real quick too.
- The Cloud vs. Local Servers for Enterprise Systems: The big advantage for the cloud here is in the ability to expand storage as needed. However, enterprise systems that really lag on Wide Area Networks (WAN) compared to Local Area Networks (LAN) will suffer.
- The Cloud vs. Desktop PCs for Software Applications: The advantage for the cloud here is in processing. There have been and continue to be many software application issues that are limited by the computing power of the desktop. However, data transferral for desktop PCs is extremely fast compared to WAN for the cloud.
The Cloud and Engineering’s Problems
So we’ve talked about the cloud generically. But let’s get down into some of the applications of the cloud specifically for engineering IT.
- Loading Large Models (CAD & CAE): In recent years, software providers have developed new compression technology, offered configurable model representation control as well as moved to 64-bit architectures to overcome RAM limitations and application responsiveness issues. However, there’s always demand for faster responsiveness and support for larger models. Utilizing the processing power of the cloud offers some promise in this area. An outstanding question is how to get around the data transmission issues. Any model large enough to cause application responsiveness issues will take a considerable amount of time to transfer over a network. A potential solution is to have the model always be in the cloud and as a result take advantage of local resources there.
- Visualization of Large Models (CAD & CAE): This issue is similar to the problems with loading large models, but the limitation is in the graphical processing power instead of RAM. Software providers have invested heavily here also by decoupling the graphical representation of the model from the underlying geometric definition. But like the prior issue, the potential of far greater graphical processing the cloud is promising. But the same issues around network transfer persist.
- Solving Simulations or Executing Design Studies (CAE): Providers of simulation software have been working on creating and improving solver speed and effectiveness for years. But in this case, the cloud offers a significant advantage in solve times over the desktop with its access to many compute power resources. As is the case with the prior scenarios, the question becomes how to move simulation models up to the cloud and results back to the desktop.
- Storage Limitations for Design Data (PDM/PLM): Personally, I have not heard about too many issues related to the storage needs of design data for PDM or PLM. But the cloud does offer the ability to expand storage as necessary.
- Centralized Access to Design Data (PDM/PLM): With so many manufacturers working across multiple remotely located design centers, centralized access to design data has become critical. The problem has been providing quick access to this data (via LAN instead of WAN) by synchronizing multiple PDM or PLM installations at each site. Personally, I don’t see how the cloud address this problem. In fact, this might exacerbate the problem if using PDM or PLM in the cloud means that moving from a LAN to a WAN.
What’s my takeaway? I’ve come to a few conclusions.
- CAD and CAE in the cloud offers some real technical advantages. Specifically in utilizing processing power in the cloud to address load, visualize and solve large models or simulations. But the outstanding issue will be in data transfer unless everything lives in the cloud and is remotely accessed. Are we heading back to the terminal-server days?
- The technical value of PDM and PLM in the cloud for users is unclear. The flexible storage capability of the cloud is clearly advantageous, but I’m not sure it has a real benefit. And in this scenario, hosting these enterprise systems in the cloud do not address the network issues typically associated with centralized access to design data.
- But there’s more value to PDM and PLM in the cloud that technical based reasons. In particular, if it is the underpinning of a SaaS offering which can be purchased more easily on a subscription basis instead of licenses, then there’s an enabling financial advantage. Furthermore, if it enables an organization to run a smaller IT team, then there’s a total cost of ownership advantage also. So it’s not just about technical advantages.
So that’s it. Time to weigh in! What other advantages or disadvantages of the cloud are there? Are there advantages to hosting PDM or PLM in the cloud I missed? Sound off and let us know your thoughts.
Take care. Talk soon. And thanks for reading.