COVID-19 has accelerated the movement of business data to the cloud, specifically to adapt to the increased scale and scope of remote work, and focusing on new capabilities for remote workforce productivity, security and collaboration. This article reveals the results of a November 2020 cloud perceptions study commissioned by HPE. We focused on the future of cloud technology, discovered some interesting trends in the manufacturing industry.
Executive summary
One of the changes triggered by COVID-19 has been an acceleration of moving data to the cloud, focused on increasing the scale and scope of remote work, and specifically adding capabilities for remote workforce productivity, security and collaboration.
This article reveals the results of a November 2020 cloud perceptions study commissioned by HPE.
The survey was focused on uncovering evolving attitudes towards the cloud, and captured 4,555 responses from IT decision makers, line of business (LOB) managers, developers, and data scientists across North America, Europe, India, Japan and Asia-Pacific (APAC) regions, and across Industries.
We focused on the future of cloud technology, and are happy to share some of our key discoveries in the manufacturing industry.
Key findings summary
- The majority of manufacturing respondents believe that public cloud reduces their overall burden and enables accessibility, openness, and speed. However, respondents are beginning to realize cost drawbacks, and pushback in some areas including security, TCO, and on what the future of hybrid cloud will look like.
- Public cloud fans were more likely to be in Japan and North America, and relatively less likely to be in India and APAC.
- For workloads, AI/ML, VDI, and remote collaboration represent the biggest move to on-premises for manufacturers – but this varies widely by geography.
Many companies are wasting about 15% of their IT investments due to over-provisioning, both public and on-premises, which needs to be eliminated.
Methodology
For manufacturing, 1,295 respondents shared their opinions on the benefits and risks of the public cloud, as below:
Participants 
Survey Questions
Respondents were asked to rank, over a set of multiple questions that you see here, in terms of “what do you believe to be most important?” These were a series of “forced choice” questions – meaning, participants had to pick one answer or the other.
The chart below represents the results of the forced choice exercise for the manufacturing industry. Anything in purple represents the answers that lean more toward a public cloud preference, while those in green represent a preference for on-premises.
For example, on the far left you’ll see a column marked “Data Accessibility”. Only 25% of respondents here report that they believe that the public cloud makes their company’s data “less accessible due to data egress costs”, while 75% believe that the public cloud allows their data to be “more accessible" from anywhere.
As you can see, the vast majority of manufacturing respondents believe that public cloud reduces their overall burden and enables accessibility, openness, and speeds their digital transformation journey, by freeing them from the burden of managing day to day resources. However, respondents are beginning to realize cost drawbacks and pushback in some areas including security, TCO and on what the future of hybrid IT looks like.
Results by geography
The chart below shows the data by geography. As we stated in our summary, public cloud fans were more likely to be in Japan and North America, and were relatively less likely to be in India and APAC.
Workload shifts towards on-premises
The chart below shows current location of workloads. A third or more of manufacturers put management, AI ops, and data mining workloads all – or mostly all – on-premises.
Workloads are projected to move slightly away from public cloud, and two in five manufacturers cite App Dev/Testing and VDI as top workloads on-premises in the future, over infrastructure management.
The next chart shows the projected location of workloads in the future (two year window), and also demonstrates that many deployments will mostly remain a form of hybrid cloud.
From the above two charts, we have created a graph showing the shift to on-premises in the manufacturing industry.
Below is the data for workload shifts by geography. AI/ML, VDI, and remote collaboration represent the biggest move to on-premises for manufacturers – but this varies widely by geography.
We’ll find out further why this move is happening in the following section.
Future needs point to real-time, low latency, and high band width on-premises
While the public cloud appeals to many manufacturers, shifts to gathering Industrial IoT (IIOT) are based on real-time insights and high performance computing requiring on-premises solutions. The IIOT revolution will lead to vast amounts of data being generated outside the data center, at the edge. Sending all data to the public cloud and back will present performance, cost, and security challenges. That's where edge computing plays such an important role – processing data at the edge. Edge computing brings real-time analytical power to the manufacturers for automation, prediction, and control.
Video analytics is also a good demonstration of the value of edge computing, as processing video near the camera eliminates the need to transfer large amounts of video data into the data center or cloud – thereby reducing bandwidth costs, greatly accelerating reaction time, and lowering the risk of corruption or espionage. Sending every bit of video data from the thousands of cameras at the edge back to a data center or cloud for processing is often too slow and expensive.
Another use case is high performance computing, or HPC. HPC provides massive computing power, along with modeling and simulation capabilities, to turn complex data into digital models that help researchers and engineers understand what something will look like, and how it will perform in the real world. HPC also provides optimal performance to run AI and analytics to increase predictability. These combined capabilities are used to solve challenges from forecasting, to improving designs of cars, planes, and even personal and consumer products. Many customers told us they are planning on-premises HPC because of “data security concerns”, “performance concerns”, and “cost of data movement”.
There are many other use cases which requires on-premises solutions, particularly at the edge. (See chart below.) This chart illustrates the target applications for edge computing. The first step is simply to use edge devices to automate the collection and pre-processing of large amounts of data; the next levels of sophistication are in the areas of response, prediction, and autonomy. These next level use cases in the chart are AI/ML, AR/VR, and 5G MEC at the Edge.
HPE GreenLake: Empowering transformation to meet digital demands
Is there a way to enjoy the same benefits and experiences of the cloud while being on-premises?
HPE GreenLake for Manufacturing is the answer. HPE GreenLake offers the best of both worlds, providing on-premises IT infrastructure – delivered with the flexibility and economics of the public cloud.
- Pay-per-use
- Scale up and down
- Simplified IT
- Managed for you
HPE has recently announced market-leading HPC solutions as fully managed, pay-per-use services through HPE GreenLake to harness data growth faster and accelerate time-to-insight.
Case study and announcement in the manufacturing industry
Toyota Mapmaster used HPE GreenLake to simplify infrastructure by integrating two systems with high-performance storage at the core, achieving cost leveling throughout the lifecycle by utilizing an IT consumption model when introducing equipment.
HPE will provide the crucial artificial intelligence (AI) and high-performance computing (HPC) infrastructure Zenuity needs to develop next generation autonomous driving (AD) systems. The end-to-end IT infrastructure will be delivered as-a-Service through HPE GreenLake.
Porsche Informatik requires operational consistency for its hybrid cloud infrastructure in order to provide digital tools to approximately 30,000 employees and a global network of automotive dealers. HPE provides Porsche Informatik the flexibility to consume hybrid cloud computing power as-a-Service, supporting a high degree of automation efficiency combined with state-of-the-art security.
Conclusion
With HPE primary research, we found that the public cloud appeals to many manufacturers – but some are starting to look at other alternative solutions because of TCO and security concerns. Even for applications and processes that require real-time, low latency, high band width and do not suit a public cloud solution, there are other solutions which provide the simplicity, agility, and economics of public cloud, as well as the security and performance benefits of on-premises IT.
You can download the full report @ edge.nxt.
Meet HPE Community blogger Kohei Kubo, who is the EMEA & APJ Industry and Workload Marketing Manager. He is responsible for driving edge to cloud solutions in the Industry and Workload Marketing team. Kohei Kubo is based in Tokyo, Japan.
Connect with Kohei Kubo on Twitter and LinkedIn!
TechExperts
Our team of HPE and other technology experts shares insights about relevant topics related to artificial intelligence, data analytics, IoT, and telco.