THE MANUFACTURING INDUSTRY CHANGING WITH INDUSTRY 4.0

With Industry 4.0, with smart connected products and factories where they are made, we are on the verge of unifying automation and artificial intelligence. This means we can leverage the full range of new technologies with a single platform:

• Enable employees in different locations to collaborate securely and in real time during working hours. Engineering, manufacturing, architecture, education, and healthcare workers often have large, graphics-intensive data sets that need to be processed securely and in real-time in 3D.
• Outside of regular working hours, the same infrastructure can be used to solve complex, compute-intensive tasks. Large enterprises with offices in different time zones can move their work from time zone to time zone using the same infrastructure. Data remains secure in the data center or cloud. Network speed and latencies are not affected.

Traditional workstations are expensive. the cost of a workstation is significant. A powerful power supply, expensive CPUs, expensive memory, local SSD/HDD storage, a dedicated GPU - all these are significant cost factors. But that's just the beginning - The administrative overhead for software and hardware maintenance is also exorbitant. Imagine your admin or admins running from one workstation to another in a far-flung building, correcting errors, installing new CAD software, replacing broken data disks, restoring data, and their designers twiddling their thumbs. Further costs are incurred when a lot of data has to be transferred or restored from central storage to the workstation over the often slow network.

Traditional workstations sit unused most of the time. when working hours are over, they only consume power because they are rarely switched off. In contrast, virtual workstations can continue to work at night. Global enterprises can have their work move from time zone to time zone on the same infrastructure with the sun. Other companies solve AI- or compute-intensive tasks at night, and the day belongs to design and knowledge workers.

Traditional workstations are insecure. because important data is distributed throughout the enterprise and often stored multiple times, data can be easily corrupted. Data can also be easily transferred to foreign media and stolen. When employees visit clients or work on shared projects, confidential, important data is often stored on heavy, expensive notebooks. They don't want to take them with them on their lunch break. Data theft or the theft of entire PCs is said to have occurred.

There is a lot to be said for a virtualized IT infrastructure. Before starting a virtualization project, the goals should be clearly formulated.
Common reasons for this are

Connect mobile and local employees?

• Preventing decentralized data storage?
• Standardization of the infrastructure?
• Low process and workplace costs?
• Work and location independence?

If you recognize yourself here, a virtualization project could be the right path for your company. To determine the virtualization needs of your IT infrastructure, we offer different approaches, from a remote demo environment in our data center in Bremen to an on-site proof of concept (PoC).
But now let's come to the pure cost comparison between physical and virtualized workstations - Here everything speaks clearly for virtualized workstations.

• Optimized license usage (distribution according to need/cost).
• Cost savings due to longer usage times of the end devices
• User connection within seconds
• Centrality - data management - data security - updates
• Mobile workstations (home office, office workstations via inexpensive end devices - the server with the virtual workstations means that the data always remains in the data center or in the cloud).
• Optimization of the loading time (10 to 400 Gbit/s)
• Reduction of administrative effort many times over through central updates, etc.
• Use of new AI techniques, such as virtual twins, pattern recognition, etc. - essential for business survival.
• Parallel operation possible - optimization of the user experience
• Load time optimization (10Gbit)
• The extreme performance of GPU processors depends on high data supply with low latency. Current memory systems cannot achieve this. when investing in new storage technology for the AI sector, please make sure that only SSDs are used. In many cases, these will be NVMe SSDs that exchange data directly with the data storage bypassing the slow PCIe bus. Network connections should preferably be connected via Infiniband (IB), as this combines the highest transfer rates with the lowest latency. This way, you get the most out of your infrastructure. High-performance and hard disk drives should go their separate ways from now on.
• Cost savings by using servers for compute-intensive tasks (rendering, AI) outside of regular working hours. internationally active companies could move their work from time zone to time zone and thus carry out developments cost-effectively around the clock on one platform and be the first to market with new products.