Technology has brought great gains to enterprise while also upping the pressure. Companies can do more, faster, with comprehensive, integrated software platforms and powerful hardware architectures, but they can’t rest on their laurels. The game is always changing. There’s always a next, new, faster, more powerful tool to buy if they want to stay ahead of the competition.
The additional challenge comes from margins being squeezed throughout the manufacturing process. Customers demand more information, more options, and more speed in product delivery. Manufacturers play a constant game of “keep ahead of the Joneses,” while trying to achieve maximum ROI.
The Purchase-and-Use Approach to Hardware
A corporation’s typical approach to computing is to equip and staff a full data center, purchasing racks of servers and piles of software to house everything from databases to corporate email, VOIP, file storage, department files, and enterprise applications. Every data center also requires support: space, power and cooling, switches, UPS, firewalls, cabling, racks, and fire protection—not to mention the staff to run and maintain the equipment. And upgrades every three to five years. Then there are the hidden costs, including outages, maintenance, backups, depreciation, insurance, and more. It’s a resource-intensive endeavor, but to date, there’s been little choice.
Hardware is a particularly tricky part of the equation to consider, especially when an enterprise needs high powered computing (HPC) configurations. HPC systems are vital for a variety of critical computer-intensive simulation and optimization tasks, such as analyses of fluid flow dynamics, mechanical and structural stresses, and electromagnetic behavior. As might be expected, the speed and capacity of installed HPC resources can have a big impact on how quickly a new product makes it to market and how it fares against the competition.
But it’s not as easy as “the bigger the better.” HPC resources are expensive, and only rarely in a typical work cycle are they used to their maximum capacity. But companies can’t skimp on them either. IDC reports that hardware and software were only 14% of the three-year cost of ownership of HPC resources, but staffing is more than half. With the greatest investment coming in the form of worker time, downtime and lost productivity that result from workers waiting around for computations to run can be the most costly gamble of all.
The Cloud Solution
In a new paper, Cloud-Based High Performance Computing, Dassault Systèmes suggests a new solution that scales with an enterprise’s needs: cloud-native software with integrated simulation tools—the 3DEXPERIENCE platform—to run on cloud HPC or in a hybrid cloud/in-house configuration.
To date, cloud hasn’t been widely adopted by those needing HPC, in part because every solver algorithm may perform better on different types of architecture, but also due to concerns about security for IP-related data, full access to technology, costs, and more. However, the cloud HPC market has risen to those challenges, much as other SaaS cloud providers did. Since Salesforce’s pioneering effort with CRM in cloud services, other enterprise cloud-service businesses have grown exponentially in recent years, from basic office tools to big-data analytics providers.
As the paper points out, Dassault is in a unique industry position with their offer of a complete suite of connected tools, “all the way from early stage product conception, through physics-based optimization to manufacturing simulation and in-service operation.” In addition, they offer an all-in-one, turnkey solution for cloud-based simulation that includes software, data, hardware, and licensing on-demand. Like today’s cloud-based CRM, cloud-based engineering and design data can be selectively shared and collaborated on by worldwide teams while still maintaining a secure, traceable single source of truth.
With Dassault’s 3DEXPERIENCE and cloud offerings, large enterprises can build confidence in cloud HPC while still working in a hybrid manner with their own datacenter, and small companies can maintain or move to a zero-IT footprint. For every organization, a natively cloud-based, collaborative system offers faster time-to-market and lower costs via improvements in product organization and management, communication, and overall productivity.
To find out more about the details and costs of HPC and Dassault’s solutions, read the paper or contact Adaptive.