Darnell Defines the Next Decade – Software-Defined Power Impacts Electronic Systems
Digital power is well established and software-defined power architectures (SDPAs) will build on that foundation. Concepts based on software-defined networks (SDNs) are beginning to emerge and be refined. SDPAs will result from a merging of developments related to SDNs and today's digital power technologies. With improved sensing and measurement technologies, more operational data is becoming available in real time. At the same time, more processing power (for lower monetary cost and lower energy cost) is increasing the sophistication and complexity of monitoring and control software.
Those trends will enable the development of software-defined power: real time and simultaneous optimization of power distribution, power conversion, energy storage and power consumption on both a global and granular basis. SDPAs are a multi-faceted, generic concept that will have broad and varied implications. This concept is independent of scale. Today we will consider the implications of SDPAs on 5G telephony and high-performance computing. Tomorrow, we will dive into the implications of SDPAs for advanced microgrids and the smart grid.
In July of last year PowerPulse reported on the start of the first phase projects of the European 5G Infrastructure Public Private Partnership (5G-PPP). This is a major milestone in a joint initiative between the European Commission and the European ICT industry to get research investment focused in a very coherent way on the challenges associated with having a communications infrastructure capable of coping with all future demands by 2020. A major goal of this effort is to achieve a saving up to 90% of energy per service provided. The main focus will be in mobile communication networks where the dominating energy consumption comes from the radio access network. A final standard is expected sometime in 2018. Field trails are expected to start in 2016. And the 2018 Olympics in Korea are expected to be the first wide-scale public demonstration of 5G.
"One of the most urgent and critical challenges in the design of 5G wireless communication systems is that of mitigating their energy consumption in light of a greatly increased data rate. Indeed while it has now been some years that the topic of energy-efficient design of communication networks has gained a prominent role, due to both economic reasons and environmental concerns on sustainable growth, it is expected that in the 5G era, with millions more base stations and billions of connected devices, this topic will be even more compelling. Given the required 1000x increase in offered data rates and throughput with respect to current state-of-the art technology, the network energy efficiency must be improved by at least the same factor, in order to keep the energy consumption at today’s level, " commented Dr. Alessio Zappone with TU Dresden.
"While software is playing an increasingly prominent role in networking at a system architecture level, it also has growing importance for the delivery of power in the future. The use of power-optimization software algorithms and the concept of the SDPA are all being seen as part of a brave new future for advanced board-power management. Energy management at the board level has evolved over the past few years, starting from virtually nowhere a decade ago to the latest advanced power systems that make use of digital monitoring and control software to improve conversion efficiency - SDPA is a natural evolution of digital power" Patrick Le Fèvre, formerly Marketing & Communication Director, Ericsson – Power Modules, commented at the time of the announcement.
In October, 2015, PowerPulse reported that the IEEE has formed the IEEE P2415â„¢ Unified Hardware Abstraction and Layer Working Group and IEEE P2416â„¢ Power Modeling Meta-standard Working Group. Both new working groups will be drafting standards that are intended to support development of more powerful and economically affordable electronics. And power-aware software is expected to be a key building block for future software-defined power architectures.
The IEEE P2416 "Standard Project for Power Modeling to Enable System Level Analysis" is being created to propose a meta-standard focused on parameterization and abstraction, enabling system, software and hardware IP-centric power analysis and optimization. This standard, once completed and approved, will define the required concepts and semantics for the development of parameterized, accurate, efficient and complete power models for systems and hardware IP blocks usable for system power analysis and optimization. These concepts include process, voltage and temperature (PVT) independence, power and thermal management interface, workload and architecture parameterization. Resulting models will be suitable for use in both software-development and hardware-design flows, as well as in representing both pre-silicon estimated and post-silicon measured data.
Also in October, PowerPulse reported that the concept of Software Defined Batteries (SDBs) for next-generation computing systems is described in a paper that a team of researchers from Microsoft Corp, Tesla Motors, University of Massachusetts Amherst, and Columbia University will present at the ACM Symposium on Operating Systems Principles this week. Different battery chemistries perform better on different axes, such as energy density, cost, peak power, recharge time, longevity, and efficiency. Mobile system designers are constrained by existing technology, and are forced to select a single chemistry that best meets their diverse needs, thereby compromising other desirable features. In this paper, the authors present a new hardware-software system, called Software Defined Battery, which allows system designers to integrate batteries of different chemistries.
SDB exposes APIs to the operating system to control the amount of charge flowing in and out of each battery, enabling it to dynamically trade one battery property for another depending on application and/or user needs. Using microbenchmarks from the prototype SDB implementation, and through detailed simulations, the authors’ demonstrate that it is possible to combine batteries which individually excel along different axes to deliver an enhanced collective performance when compared to traditional battery packs. SDB allows a device to use diverse batteries through fine-grained control of the amount of charge flowing in and out of each battery. SDB provides APIs to the OS to change the power values based on user workload.
In May, Microchip Technology Inc. announced the 14-member dsPIC33EP "GS" family of Digital Signal Controllers (DSCs). The dsPIC33EP "GS" family delivers the performance needed to implement more sophisticated non-linear, predictive and adaptive control algorithms at higher switching frequencies. These advanced algorithms enable power supply designs that are more energy efficient and have better power supply specifications. Higher switching frequencies enable the development of physically smaller power supplies that offer higher densities and lower costs. Compared with the previous generation of DSCs, the new dsPIC33EP "GS" devices provide less than half the latency, when used in a three-pole three-zero compensator, and consume up to 80% less power in any application.
"Today’s systems are more energy efficient than ever before. The improvements have come from a combination of lower-power components, such as application processors that offer more performance per watt of power used, and from more advanced power-conversion stages that effectively deliver power to the loads. Using digital-control technology, the overall efficiency of ac-dc power supplies has been improved by implementing adaptive techniques—such as phase shedding, dead-time adjustment, variable switching frequency, variable bulk voltages and other practices—that maximize the efficiency over widely varying load conditions. Extending this concept of adaptive techniques even further, additional energy-efficiency improvements at the system level can be brought about through a software defined power architecture (SDPA), as Jeff from Darnell Group has described, " commented Tom Spohrer, Power Product Marketing Manager, MCU16 Division, Microchip Technology Inc.
Digital power is well established and software-defined power architectures (SDPAs) will build on that foundation. Concepts based on software-defined networks (SDNs) are beginning to emerge and be refined. SDPAs will result from a merging of developments related to SDNs and today’s digital power technologies. With improved sensing and measurement technologies, more operational data is becoming available in real time. At the same time, more processing power (for lower monetary cost and lower energy cost) is increasing the sophistication and complexity of monitoring and control software. Those trends will enable the development of software-defined power
