Electronics Product Innovation in the Era of Software-Defined Everything and Virtualization
Disruption Deep Dive
Overview
Hardware is getting standardized and commoditized, from everyday consumer devices to data center servers, from special industrial automation assets to telco equipment. Development of future electronics product or equipment is increasingly dependent on delivery of virtualized device functions and value-added by software on off-the-shelf hardware, to reduce cost, add flexibility and allow portability. To achieve this, electronics companies will need to migrate more development funding to software, services and surrounding ecosystem ventures to change the shape and value of electronics organizations. Virtualization will also impact how electronics companies collaborate with their customers and partners, such as cloud service providers or telco operators, to clearly define the value creation and boundaries of responsibilities. We believe that cloudification with virtual machine and container technologies will help drive this trend, fundamentally changing the way electronics device functionality is delivered, and also leading to win-win situation in the electronics value creation ecosystem.
In this article, Peter Xu, IBM Electronics CTO and Martin Kienzle, Electronics Industry Leader at IBM Research, walk you through definitions, technologies and industry-specific recommendations.
Virtualization: Context-dependent
Virtualization — one word, dozens of potential meanings. When talking about virtualization, it’s really important to get the context right based on the stakeholders and industry segment you’re talking to.
For instance, in the IT vendor world, virtualization refers to the act of creating virtual versions of computing resources, including servers with CPU and memory, storage devices and computer network gears. Increasingly, application agility and portability require new forms of virtualization.
For communication network equipment providers, it might mean either Radio Access Network Virtualization (vRAN) or Network Function Virtualization (NFV). Virtual RAN enables operators to run baseband unit (BBU) functions on off-the-shelf hardware as virtual network functions.
For consumer electronics and industrial device manufacturers, Device Function Virtualization (DFV) is most relevant and often discussed. It may imply that critical device functions are increasingly implemented with software, which could even be moved to a different computing host.
We will dive deeper into these topics and see how virtualization technologies have revolutionized or will fundamentally change the way electronics companies deliver differentiated products and services.
Hardware and software virtualization
Hardware virtualization or platform virtualization “creates” a virtual machine (VM) that acts like a real computer with an operating system. Software executed on these virtual machines is separated from the underlying hardware resources. This enables sharing the hardware, driving resource utilization. That, in turn, enables sizable cost reduction for the users. Cloud computing is the poster child of the great benefits virtualization technology can bring. You can host multiple VMs in one physical PC, or you can maintain hundreds of VMs across a pool of servers. In the telco equipment space, mobile carriers spend most on RAN base band devices. Virtualizing some of those device functions allows an operator to process signals on standardized hardware in data centers, and can lead to major cost savings.
But this comes at a cost to hardware providers: standardization also means commoditization, meaning lower prices and less ability to charge premiums. Hardware vendors are all pressed to find new sources of differentiation in software and services. As software is taking over the world, electronics companies are increasingly embedding more software-based functions into their products: some products disappear, replaced completely by software; some products move away from proprietary hardware, and are redesigned based on custom software and standard off-the-shelf hardware; some products add IoT connectivity or embed AI capabilities.
Traditionally, software-intensive system engineering, packaging and deployment is complex and time-consuming. This is due to an inherent software dependency: you have to consider what operating system you’re using, the resources the applications require such as specific system drivers for sound, networking or graphics cards, or the specific version of SDKs for frameworks your applications depend on. You also need to decide where and how it should be installed. All of those above often lead to monolithic applications which are hard and costly to debug, change and port. This is especially true for software-intensive system and product engineering, so it is not surprising that we see increasingly higher percentage of software cost as part of overall product development cost.
Enter the world of containers and Docker
Fortunately, over the last couple of years, we are seeing the maturing and wider adoption of a container technology called Docker, which already has made major impact on software and system engineering. It provides a level of abstraction for software packaging and movement. You can think of a Docker container as a physical shipping container. It’s a box where you store and run an application and all of its dependencies. Just as cranes, trucks, trains, and ships can easily work with shipping containers, so can Docker run, copy, and distribute containers with ease. Docker completes the traditional container metaphor by including a way to package and distribute software.
The component that fills the shipping container role is called an image. A Docker image is a bundled snapshot of all the files that should be available to a program running inside a container. You can create as many containers from an image as you want. But when you do, containers that were started from the same image don’t share changes to their file system. When you distribute software with Docker, you distribute these images, and the receiving computers/system create containers from them. Images are the shippable units in the Docker ecosystem. This self-contained nature and portability are the key to software function abstraction and movement. This new portability makes it feasible to run the same software — exactly the same software — on any system. That means your desktop, your development environment, your company’s server, and your company’s cloud can all run the same programs, as long as physical resource capacity allows it. The same mobility applies to software running inside electronic devices. This is a huge advantage for the embedded development community. For the first time, it is possible port device capabilities easily and with confidence to gateways, fog computing environment, cloud, or wherever it makes sense, essentially “virtualizing” device functions.
How electronics companies are adopting and innovating with virtualization technologies
There is no single pattern fitting all cases, since electronics products diverge greatly: from semiconductors to consumer electronics, from PCs to data center servers, from automation and power equipment to telco and communication equipment.
As we discussed above, hardware virtualization forces electronics companies to look for ways to differentiate. We see increased R&D spending for special chip design including AI chips, specialty chips and FPGAs. At the same time, more electronics manufacturers are using a combined hardware-data-experience strategy that sells hardware (H), analyzes data (D), and then uses the hardware to improve client experience (X). Areas of innovation include:
- Connect and manage the IoT products
- Embed software, especially analytics and AI into products from the ground up
- Add intuitive user interactions such as voice, visuals and gestures to the product
- Build services around the product that share product usage and sensor data and build the platform for ecosystem partners. see the IBV’s Plotting the Platform Payout
- Outsource non-core functions or software development
That’s a tall order for traditional electronics companies. Fortunately, as we elaborated above, adopting software virtualization technologies such as Docker will enable flexible device function creation and movement. Software defined device functions, i.e., the use of containers for resource abstraction enables the development and wide-spread use of functions, on pretty much any platform. One final note here: don’t try to do it all yourself. Ecosystems play a crucial part in developing and extending capabilities that should not be overlooked.
Virtualization in Action
This is the Software-Defined-Everything world we live in, and we see how leading electronics companies, working with their partners, are adopting virtualization as the way to innovate and transform.
For traditional IT hardware and cloud server vendors, decreasing hardware margins force them to add more value in software, ranging from automated deployment to full life-cycle management with virtual machine technologies. Some even go further to add native Docker and Kubernetes orchestration support, to allow easier provisioning of any software package. This opens the door for deeper integration of streaming analytics and machine learning workloads from third parties. Scalable and fully integrated software-hardware appliances will give the end customer turnkey solutions, saving months of work stringing pieces together. Toshiba Flashmatrix® is such an example: it fully virtualizes storage in a shared everything architecture and integrates a converged mesh network switch in a 2U enclosure, and includes Daisho Core, a web-based management interface to manage all compute, storage and networking resources, and supports the super orchestration of Docker® container-based microservices.
In Telco, mobile operators are looking for new ways to cope with ever-increasing data traffic while improving the operational and capital efficiency of their networks. Cloud computing with its virtualization and network function virtualization (NFV) have emerged as key enablers to optimize resource utilization and at the same time reduce network operational expenditure (OPEX). In virtualized networks, network functions are delivered as software running on generic hardware, allowing service providers to dynamically allocate resources based on traffic and service demand. In addition, telco equipment providers have been testing virtualization technologies for the emerging 5G networks, which promise increased speed and reliability, lower latency, and more bandwidth. For 5G to be successful, its backhaul operation needs to be very flexible. Software defined networking (SDN) and network function virtualization (NFV) are the technologies embraced to support this. In some implementations, VMs and Docker containers have been used to prototype software defined Radio Access Network (vRAN). Results have verified the superiority of the Docker technology. Further research is focused on the feasibility of development of container-based 5G SDN controller. The telco equipment provider Nokia has been working hard on this, and its AirScale Cloud RAN is a great example of taking advantage of software and cloud to virtualize some hardware functions.
In the IoT gateway space, we have seen innovative products like MICA from Harting.
MICA (Modular Industry Computing Architecture) provides a quick and easy solution for implementing digitization projects directly on factory equipment. With its modular open platform, MICA can be customized with hardware, software and interfaces — to suit individual requirements for different industries. Its Virtual Industrial Computing allows users to simultaneously operate multiple sensors, field devices or processes independently in separate containers and isolated from one another. If an application changes, only the corresponding container is affected. The rest of the system continues.
We live in a truly exciting software-defined-everything world. Because of virtualization and cloudification, we see information technology (IT) converging or disrupting communication technology (CT) and operation technology (OT) world. Electronics product companies are at the center of this storm. Those that can adapt at speed of changes will be reaping the rewards of innovation.
Contact the authors, Peter Xu and Martin Kienzle
In or Out: Succeeding in the New Ecosystem Economy
https://www-01.ibm.com/common/ssi/cgi-bin/ssialias?htmlfid=GBE03850USEN&
