10 Nov Cloud CFD Advantages – Part 4 of 4: Computing Hardware
There is an unstoppable trend in the CFD software industry: we are all moving on the Cloud. This process has been happening in a lot of contexts, some of them similar to CFD (like CAE in general) and some quite different (software management, PLM, CRM, etc). It is considered the next future by a lot of experts.
In the software sector it basically means a change from the Software-as-a-Product (SaaP) to the Software-as-a-Service (SaaS) paradigm. This epochal change is a huge opportunity for both software vendors and providers as much as for software users. There are a lot of advantages that can be found, and especially in the CFD field, some of them are quite relevant.
We grouped them in four categories: Financial, User experience, Accessibility, Hardware. After the first post about financial aspects, the second one about the user experience and the third about accessibility opportunities, this post will briefly review the fourth of the series: the computing hardware.
One of the most effective choices to deal with high workload peaks is to rely on parallelization. To perform multiple tasks at the same time allow to ideally reduce the total amount of time needed, down to the longest single independent task.
Engineering Simulations are a typical example of tasks that can benefit from such an approach. Especially when designing a new product, multiple slightly different simulations are needed to investigate all available options. All these jobs can be run in parallel, and all results will be available at the same time, waiting just the time needed for a single simulation.
In order to take advantage of parallelization, the “old-fashioned” approach of owning powerful computational hardware soon reveals itself as not profitable. In fact running n simulations in parallel would require n-times the computational power needed for a single simulation and at the same time would then leave those expensive machines unused for the ((n-1)/n)*100 % of the time.
With modern Cloud Computing solutions, both problems are easily tackled. As shown in the image on the right, with CONSELF you can run as many simulations as you want at the same time, directly from your ordinary laptop browser. You don’t need to buy high performance hardware anymore but you can rely on unlimited computing power just when you need it.
Time-to-market is central in modern manufacturing. Speeding up design process is usually related with an increased production cost that can then cause higher end prices or lower margins.
With simulations performed via old standards, where software products licences are sold and they are run on high performance workstations, the above mentioned problem affects engineering analysis too. Something completely different happens when innovative simulation products are considered.
These two different contexts are described in the image on the left, where a company is supposed to run a simulation that requires a given amount of “computational time”, quantified as the product of number of CPUs times the hours they are running. Under the assumption of linear scalability, the solution time (time needed to obtain the solution) is inversely proportional to the number of CPUs used. But a major difference is then encountered between old simulation products and new, Software-as-a-Service, ones: while for the former costs can be approximated as directly proportional to the number of CPUs, for the latter costs are directly proportional to the computational time needed (CPUs x time), that can be approximated as constant. It is then clear how software as CONSELF, where you pay only the computational time you need, allow to dramatically reduce solution time while maintaining costs constant.
Another important aspect when dealing with hardware resources is maintenance, that is a relevant cost. When a powerful infrastructure has to be maintained it needs qualified personnel to assure its constant availability. In addition breakdowns problems must be addressed, assuring countermeasures to avoid all of them, even those not directly related with the infrastructure itself but with its connected elements (e.g. power supply).
Relying on Cloud based solutions as CONSELF allow to save all burden related with privately owned infrastructures. At the end of the road it is all a matter of costs, and the strength of the software-as-a-service approach is mainly linked with this fundamental aspect: it lowers costs/investments up to 90% with respect to traditional solutions.
CONSELF wants to make state of the art, cutting edge technology, available to every professional in the globe.
CFD & FEM simulation software is a very powerful tool, with its adoption optimization and innovation can be achieved in every field. To make this instrument accessible to everyone means lowering costs, but mainly to develop an infrastructure that favours a super-easy adoption by market new entrants.
CONSELF is highly committed and will pursue this goal working side by side with professional and industries to define the best strategies and solutions.
Alessandro Palmas – firstname.lastname@example.org