Evolving Design Software Underscores Importance for Right Hardware
Articles • Central Innovation • 27 February 2017
The rapid evolution of design software is equipping manufacturing sector businesses with precision tools of ever greater sophistication. With the increasing complexity of these software systems comes greater demand on the corresponding hardware platforms – for example, the processing power of the workstations on which CAD software packages are run.
At the same time, we are entering a new era of increased automation and digitization, with robotics and Virtual Reality set to become more commonplace, and the demands on the software necessary for the implementation of these technologies again underscores the importance of having the right hardware for the job.
Jason Princehorn, Manager, Hardware Solutions for Central Innovation offers some insights on the subject along with a few tips which might save you unnecessary expense.
“In hardware terms, your typical workstation has a lifespan of three to five years,” Jason points out. “So the question you need to ask is: what is the right hardware that I need, in order to keep up with the evolution in my software over that period?”
While one answer might be to invest in the most expensive workstation you can afford, or the one with the highest processing power, by doing this you could be spending more than necessary.
“When an existing or prospective customer comes to us, we don’t ask them what computer hardware or what specification script they want,” Jason explains. “Instead, we ask what they’re doing – what software they’re using.”
By identifying what applications the software is used for, and the workflow requirements around this, it’s then possible to create a bespoke workstation that will deliver optimal productivity.
“For example, if you’re doing design simulations and rendering, how much time are you sitting in front of the computer each month while it’s rendering, and is that equating to lost productivity? If so it can be worth investing in additional CPU power to save you time. But sometimes customers come to us wanting higher processing power and we’re able to steer them in another direction, which in the end costs them less and gives them a faster solution for the task at hand.
“You might think that because one workstation costs more, it must perform better – but that’s not necessarily the case. For example if you’re using the base modelling package of Solidworks, it’s predominantly a single-threaded application, which means what’s really important is to have the fastest CPU cores you can get.
“So you need to understand what’s required to get the job done, and to do that sometimes you need to get expert advice from someone who can explain everything from the ground up – rather than sell you something that won’t actually save you money and time.”
As an example, Jason cites the evolution in graphics card technology: “Over the last two generations of Nvidia GPUs (Graphics Processing Units), processing power has at least doubled for any given level of GPU. So whereas in the past you may have needed a high end graphics card to work effectively on large and or complex assemblies, today this is no longer the case and you may be able to get away with a graphics card that costs half as much as your previous one, yet still provides more graphics processing power.
“However there are other considerations to take into account – such as how many displays are you likely to have in three years’ time and what resolution will they need to be? The availability of quality 4K screens will bring with it the need to ensure your graphics card can still effectively run the software for your projects.
“So it’s not simply a question of whether according to the manufacturer’s specifications the software can run this number of monitors at this resolution – it’s also the fact that you need to be happy with the overall experience of what you see on screen.”
Jason argues that the key in ensuring the right hardware for the job is to understand how the software interacts with the hardware – what parts of the software place demands on what parts of the hardware, and how they do it.
“Is it CPU based or GPU based, how much memory is involved, what type and speed of drives are most beneficial to what type of workflow? And most importantly, do I need more CPU cores or faster CPU speed? So you need to understand all the factors involved to build a workstation that will give you the best ROI for your workflow.”
Jason offers a further example: “With simulation becoming more and more multithreaded, you may think that getting a high end Dual Xeon system is going to give you the fastest simulation workstation. However in most of the scenarios we look at, this is actually not the case as adding additional cores from an off-the-shelf solution usually means much lower CPU speed and ends up with greatly diminished returns.”
So when considering your future software options, it’s imperative to identify the demands that they will place upon your hardware, along with the way your workflow is likely to impact on your CPU requirements. Armed with this knowledge, you’ll then be able to ensure you have the right hardware for the job.