<p dir="ltr" id="isPasted">Renewable energy sources, such as solar, wind, and hydroelectric power, play a crucial role in the global energy transition towards a more sustainable and ecological energy scenario. As the demand for renewable energy grows, so does the need for technologies that can improve the management of these resources and the transformation efficiency. One critical aspect of ensuring the viability of renewable energy projects is applying effective risk management solutions to manage generating equipment, avoiding unwanted interruptions, and optimizing system operation. In this context, single board computers (SBCs) are emerging as compact and versatile computing devices that offer technical advantages in the renewable energy sector, supporting the operation and management of renewable energies.</p><p dir="ltr">This article explores how SBCs can be used in renewable energy applications, integrating system monitoring and control.</p><h1 dir="ltr">SBCs in Renewable Energy</h1><p dir="ltr">SBCs, with their compact size and powerful hardware, offer features that make them attractive for renewable energy applications, where real-time data analysis, monitoring, and control are essential. SBCs differ from conventional computing solutions in providing flexibility, such as being installed where space may be limited presenting different connectivity options with sensors and devices. They are compatible with various operating systems, making them highly customizable and adaptable in the applications. Moreover, the SBCs offer low application costs.</p><h2 dir="ltr">System monitoring</h2><p dir="ltr">Continuous monitoring of the performance of renewable energy installations, such as solar farms or wind turbines, is essential to monitor energy generation and ensure the effective use of the system. For this purpose, SBCs communicating with various sensors can be strategically placed throughout these installations to collect data about multiple parameters, including electricity production, equipment temperatures, and weather conditions. This allows grid operators to assess the system's functioning in real-time and promptly identify potential problems.</p><h2 dir="ltr">Grid balance</h2><p dir="ltr">The integration of renewable energy into the electric grid introduces fluctuations due to the intermittent nature of these energy sources. SBCs can contribute to grid stability by optimizing the distribution of renewable energy through the data collected and running algorithms for this purpose. These devices can analyze energy supply and demand in real-time, allowing automatic adjustments to energy production from renewable sources or the activation of energy storage systems. The usage of dynamic grid management ensures a reliable energy supply, balance generation and consumption in real-time.</p><h2 dir="ltr">Predictive maintenance</h2><p dir="ltr">Conventional systems typically use periodic maintenance programs, which may sometimes be less efficient and potentially costly. In contrast, when integrated with SBCs, equipment can undergo continuous monitoring and have its data analyzed for any potential anomalies. Using machine learning algorithms and artificial intelligence, SBCs can accurately predict when maintenance is required. This proactive approach helps reduce downtime, significantly cut maintenance costs, and extend the operational lifetime of renewable energy assets.</p><h1 dir="ltr">Benefits and Challenges of SBCs in Energy Generation</h1><p dir="ltr">Due to their flexibility and powerful processing capabilities, SBCs can improve various aspects of renewable energy generation:</p><ul><li dir="ltr"><p dir="ltr">Energy reliability:&nbsp;SBCs can provide real-time insights into performance and potential equipment issues, contributing to the overall reliability.</p></li><li dir="ltr"><p dir="ltr">Rapid interventions:&nbsp;Through data analysis and instant alerts, SBCs enable rapid responses to critical events or damage, minimizing energy loss and the impact on profits. For example, overheating of equipment or anomalies in the operation of photovoltaic or wind power generation.</p></li><li dir="ltr"><p dir="ltr">Cost-effective:&nbsp;The predictive maintenance and optimized network balancing that can be done using SBCs reduce operating and maintenance costs. Furthermore, SBCs have compact sizes and low power consumption.</p></li></ul><p dir="ltr">However, there are some challenges in implementing it in renewable energy resources:</p><ul><li dir="ltr"><p dir="ltr">Initial investment:&nbsp;Implementing SBC-based systems can require significant initial investments in software and integration with existing renewable energy infrastructure.</p></li><li dir="ltr"><p dir="ltr">Technical expertise:&nbsp;Integrating sensors, data analysis software, and control systems is complex and can present technical challenges, requiring expertise in renewable energy and computing.</p></li><li dir="ltr"><p dir="ltr">Cybersecurity:&nbsp;As renewable energy systems become more connected, they are susceptible to cyber-attacks, and ensuring the security of SBC-based systems is vital.</p></li></ul><h1 dir="ltr">Case Study: SBCs for Photovoltaic Generation</h1><p dir="ltr">SBCs can be utilized in photovoltaic systems to monitor and control energy production. They can collect data from solar panels and control their operation. An SBC, such as ROCK 5B, can receive and process data to help optimize photovoltaic system efficiency and generation. ROCK 5B, developed by <a href="https://www.okdo.com/" target="_blank">OKdo</a> in collaboration with <a href="https://radxa.com/" target="_blank">Radxa</a>, is a state-of-the-art device powered by the Rockchip RK3588 SoC, Arm DynamIQ Quad Cortex-A76 @ 2.2/2.4 GHz.</p><p dir="ltr"><br></p><p dir="ltr">It features a safe shutdown on/off button and is compatible with various operating systems, including Linux and Android, for enhanced application flexibility. The ROCK 5B has extensive storage capacity, with both 8 and 16 GB of RAM options as well as a high-speed eMMC socket for eMMC modules - ideal for storing operating systems and data. Additionally, it has high compatibility with SBC accessories via the 40-pin GPIO expansion connector.</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dib" style="width: 302px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjk5MjY5NDg5NDE1LWltYWdlMi5wbmciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 300px;" class="fr-fic fr-dib"><span class="fr-inner">ROCK 5B. Image credit: OKdo.</span></span></span></p><p dir="ltr" id="isPasted">Photovoltaic systems use current and voltage sensors to monitor energy generation. Some of the popular sensors available in the market include the ACS712 for measuring electric current and the INA219 for measuring the voltage of the panel array. Real-time monitoring of photovoltaic performance is possible with these sensors, which provide valuable information for control and preventive maintenance. Temperature sensors such as PT100 and solar radiation sensors like LI-200R can be used to monitor the operating conditions of solar panels and optimize system efficiency. For example, photovoltaic generation is directly related to the solar irradiation incident on the solar panels, making irradiation an excellent metric for monitoring energy generation. The relationship between the two on a clear day is illustrated in the figure below.</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dib" style="width: 302px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjk5MjY5NTE1NTk1LWltYWdlMS5wbmciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 300px;" class="fr-fic fr-dib"><span class="fr-inner">Example of photovoltaic generation vs solar irradiation. Image credit: [9]</span></span></span></p><p dir="ltr" id="isPasted">Additionally, a web server like Apache can remotely monitor the solar energy system. The data collected and processed by the SBC can be easily accessible to interested parties, such as solar power plant operators and owners.</p><h2 dir="ltr">Other applications</h2><p dir="ltr">Wind turbine operators can utilize SBCs to monitor vibrations, temperature, and energy production continuously. With this information, the maintenance algorithms can provide valuable insights for predictive maintenance and reduce downtime.&nbsp;<iframe id="639e88944155b50008d4cdf8" width="250" class="specs-iframe" height="440" src="https://wevolver.com/iframe/specs/okdo-x-radxa-rock-5-model-b?iframe=true" frameborder="0" scrolling="false" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe></p><p dir="ltr">Another potential application of SBCs would be in smart grids, which can be employed to manage energy storage and distributed energy resources. By optimizing energy distribution and storage, this technology would enable demand response and efficient energy supply.</p><h1 dir="ltr">Future Perspectives on Using SBCs in Renewable Energy Systems</h1><p dir="ltr">As technology advances, the role of SBCs in the renewable energy sector is set for considerable expansion. With the Internet of Things (IoT) and Artificial Intelligence (AI) becoming increasingly prevalent, SBCs can serve as a promising element in increasing the efficiency and safety of renewable energy systems. The technical characteristics of SBCs position them as a powerful device for ensuring the viability and sustainability of renewable energy solutions.</p><h1 dir="ltr">Conclusion</h1><p dir="ltr">SBCs contribute to harnessing the potential of renewable energy by enabling real-time monitoring, grid balancing, and predictive maintenance. These devices offer flexibility and economy, benefiting reliability, rapid response to problems, and cost savings. Although there are challenges, SBCs show great potential for the future as technology develops, supporting more efficient and sustainable renewable energy systems.</p><h1 dir="ltr">References</h1><ol><li dir="ltr"><p dir="ltr">Philip Hirschhorn and Tom Brijs, 2021. “<a href="https://www.bcg.com/publications/2021/addressing-variable-renewable-energy-challenges" target="_blank">Rising to the Challenges of Integrating Solar and Wind at Scale</a>”</p></li><li dir="ltr"><p dir="ltr">IRENA, 2015. “<a href="https://www.irena.org/-/media/Files/IRENA/Agency/Publication/2015/IRENA_Baseload_to_Peak_2015.pdf" target="_blank" data-lf-fd-inspected-kn9eq4roawkarlvp="true">From Baseload to Peak: renewables provide a reliable solution</a>”</p></li><li dir="ltr"><p dir="ltr">Kabeyi Moses Jeremiah Barasa and Olanrewaju Oludolapo Akanni, 2022. “<a href="https://www.frontiersin.org/articles/10.3389/fenrg.2021.743114" target="_blank">Sustainable Energy Transition for Renewable and Low Carbon Grid Electricity Generation and Supply</a>”</p></li><li dir="ltr"><p dir="ltr">Johns Hopkins University, 2021. “<a href="https://energy.sais.jhu.edu/articles/the-future-of-sustainable-energy/" target="_blank">The Future of Sustainable Energy</a>”</p></li><li dir="ltr"><p dir="ltr">Hadi Ganjineh, 2021. “<a href="https://www.forbes.com/sites/forbestechcouncil/2021/11/30/how-artificial-intelligence-and-machine-learning-are-transforming-the-future-of-renewable-energy/?sh=7f961d5e541b" target="_blank">How Artificial Intelligence And Machine Learning Are Transforming The Future Of Renewable Energy</a>”</p></li><li dir="ltr"><p dir="ltr">Nicolas Sarmiento Sierra, 2023. “<a href="https://www.wevolver.com/article/enhancing-electrical-grid-management-with-emerging-sbc-technology" target="_blank">Enhancing Electrical Grid Management with emerging SBC technology</a>”</p></li><li dir="ltr"><p dir="ltr">R H P Putra and D Wahyudin and T Sucita, 2018. “<a href="https://iopscience.iop.org/article/10.1088/1757-899X/384/1/012041" target="_blank">Designing Energy and Power Monitoring System on Solar Power Plant Using Raspberry Pi</a>”</p></li><li dir="ltr"><p dir="ltr">Hamzah Eteruddin and David Setiawan and Atmam, 2020. “<a href="https://iopscience.iop.org/article/10.1088/1755-1315/469/1/012051" target="_blank">Web Based Raspberry Monitoring System Solar Energy Power Plant</a>”</p></li><li dir="ltr"><p dir="ltr">Muhammad Qamar Raza, Mithulananthan Nadarajah and Chandima Ekanayake, 2016. “<a href="https://www.researchgate.net/publication/305321699_On_recent_advances_in_PV_output_power_forecast" target="_blank">On recent advances in PV output power forecast</a>”</p></li></ol>

How single board computers (SBCs) are shaping and helping to manage and mitigate risks in renewable energy


Impact of SBCs in Renewable Energy

<p dir="ltr" id="isPasted">The new iteration features groundbreaking silicon at its heart. The <a href="https://www.mouser.com/new/raspberry-pi/raspberry-pi-5-sbc/" target="_blank">Raspberry Pi 5</a> is a game-changer with a few new tricks up its sleeve.</p><p dir="ltr">The Raspberry Pi Foundation has navigated countless challenges over the years, from production woes to supply chain disruptions. But their unwavering commitment to delivering computing innovation to all has brought us to this pivotal moment.</p><h2 dir="ltr">Introducing the Raspberry Pi 5: A Marvel of Custom Silicon</h2><p dir="ltr">The Raspberry Pi 5 is based on the Broadcom BCM2712, which houses a quad-core 2.4GHz 64-bit Arm® Cortex®-A76 processor. However, the standout component here is the RP1, a proprietary I/O controller crafted entirely in-house by the Raspberry Pi team. This custom chip assumes a pivotal role in furnishing USB 3.0, Ethernet, GPIO, MIPI display, and MIPI camera ports. The inclusion of the RP1 custom I/O chip not only adds stability to the Pi 5's manufacturing process but also optimizes power consumption by integrating only the most critical circuits into the chip.&nbsp;&nbsp;<iframe id="6567366c3a755eabfe69da44" width="250" class="specs-iframe" height="440" src="https://wevolver.com/iframe/specs/raspberry-pi-5-single-board-computer?iframe=true" frameborder="0" scrolling="no" data-gtm-yt-inspected-9="true" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe></p><p dir="ltr">But the excitement doesn't stop there. Beyond custom silicon, the Raspberry Pi 5 introduces a four-lane PCIe 2.0 interface, streamlining high-speed peripheral connections. It also boasts Bluetooth® 5.0 with Bluetooth Low Energy, dual-band Wi-Fi®, and Gigabit Ethernet, granting users the flexibility to choose between wired and wireless connectivity. Moreover, the Pi 5 extends support for Power over Ethernet (PoE), enabling users to power the board via Ethernet, alongside a USB Type-C port for alternative power sources.&nbsp;</p><p dir="ltr">While the board seamlessly integrates a real-time clock (RTC), the absence of an onboard battery necessitates using a networked time server to maintain time accuracy during power disruptions. Fortunately, users can power the RTC externally through a supercapacitor or small battery, ensuring precise timekeeping even in power outage scenarios.&nbsp;</p><p dir="ltr">Power management receives an upgrade with the Renesas DA9091 power management IC, capable of supplying up to 20A to the BCM2712 for high-performance processing. The PMIC and Raspberry Pi 27W power supply should reduce or eliminate power warnings in most scenarios.</p><h2 dir="ltr">Raspberry Pi 5 Specifications</h2><ul><li dir="ltr"><p dir="ltr">Broadcom BCM2712 2.4GHz quad-core 64-bit Arm Cortex-A76 CPU featuring cryptography extensions, 512KB per-core L2 caches, and a 2MB shared L3 cache.</p></li><li dir="ltr"><p dir="ltr">VideoCore VII GPU, supporting OpenGL ES 3.1, Vulkan 1.2</p></li><li dir="ltr"><p dir="ltr">Dual 4Kp60 HDMI display output with HDR support</p></li><li dir="ltr"><p dir="ltr">4Kp60 HEVC decoder</p></li><li dir="ltr"><p dir="ltr">LPDDR4X-4267 SDRAM (4GB and 8GB available at launch)</p></li><li dir="ltr"><p dir="ltr">Dual-band 802.11ac Wi-Fi</p></li><li dir="ltr"><p dir="ltr">Bluetooth 5.0/Bluetooth Low Energy</p></li><li dir="ltr"><p dir="ltr">microSD card slot, with support for high-speed SDR104 mode</p></li><li dir="ltr"><p dir="ltr">2 × USB 3.0 ports, supporting simultaneous 5Gbps operation</p></li><li dir="ltr"><p dir="ltr">2 × USB 2.0 ports</p></li><li dir="ltr"><p dir="ltr">Gigabit Ethernet, with PoE+ support (requires separate PoE+ HAT)</p></li><li dir="ltr"><p dir="ltr">2 × 4-lane MIPI camera/display transceivers</p></li><li dir="ltr"><p dir="ltr">PCIe 2.0 x1 interface for fast peripherals (requires separate M.2 HAT or other adapter)</p></li><li dir="ltr"><p dir="ltr">5V/5A DC power via USB Type-C, with Power Delivery support</p></li><li dir="ltr"><p dir="ltr">Raspberry Pi standard 40-pin header</p></li><li dir="ltr"><p dir="ltr">Real-time clock, powered from external battery</p></li><li dir="ltr"><p dir="ltr">Power button</p></li></ul><h2 dir="ltr">Conclusion</h2><p dir="ltr">The introduction of the Raspberry Pi 5 heralds yet another innovation from the Raspberry Pi Foundation. The inclusion of the RP1 custom I/O chip not only adds stability to the Pi 5's manufacturing process but also optimizes power consumption by integrating only the most critical circuits into the chip.&nbsp;</p><p dir="ltr"><a href="https://www.mouser.com/new/raspberry-pi/raspberry-pi-5-sbc/" target="_blank">You can pre-order the Raspberry Pi 5, from Mouser here.&nbsp;</a></p><p dir="ltr"><em>Note: This article was written by Mouser staff using OpenAI ChatGPT v4.</em></p>

The Raspberry Pi Foundation just released their latest innovation, the Raspberry Pi 5.

Meet the new Raspberry Pi 5

<p><em>This is the first in a series of articles, sponsored by Synopsys that look at AI, automotive, and multi-die chip design.&nbsp;</em></p><h2>Introduction</h2><p>Artificial intelligence (AI) accelerators are deployed in data centers and at the edge to overcome conventional von Neumann bottlenecks by rapidly processing petabytes of information. Even as Moore’s law slows, AI accelerators continue to efficiently enable key applications that many of us increasingly rely on, from ChatGPT and advanced driver assistance systems (ADAS) to smart edge devices such as cameras and sensors.</p><p>Although AI accelerators are typically 100x to 1,000x more efficient than general-purpose systems, the computational resources needed to generate best-in-class AI models <a href="https://openai.com/blog/ai-and-compute/" target="_blank" id="isPasted">doubles every 3.4 months</a>. Moreover, training a single deep-learning model such as ChatGPT’s GPT3 creates approximately <a href="https://www.theguardian.com/technology/2023/jun/08/artificial-intelligence-industry-boom-environment-toll" target="_blank">500 metric tons of CO<sub>2</sub></a>, the equivalent of over a million miles driven by an average gasoline-powered vehicle! To help reduce global carbon emissions, the U.S. Department of Energy (DoE) <a href="https://www.energy.gov/sites/default/files/2022-02/Semiconductor%20Supply%20Chain%20Report%20-%20Final.pdf" target="_blank" data-lf-fd-inspected-kn9eq4roawkarlvp="true">recently recommended</a> a 1,000x improvement in semiconductor energy efficiency.</p><p>Achieving optimal performance-per-watt—whether for AI training in the data center or inference at the edge—is understandably a top priority for the semiconductor industry. In addition to minimizing environmental impact, reducing energy consumption lowers operating costs, maximizes performance within limited power budgets, and helps mitigate thermal challenges. Read on to learn how chip designers—including edge AI chip developer <a href="https://sima.ai/" target="_blank">SiMa.ai</a>—are leveraging end-to-end power analysis solutions to build a new generation of more energy-efficient AI accelerators.</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 302px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjk1ODA4NTA3MzQ4LWNoaXAtc3VzdGFpbmFiaWxpdHkud2VicCIsImVkaXRzIjp7InJlc2l6ZSI6eyJ3aWR0aCI6OTUwLCJmaXQiOiJjb3ZlciJ9fX0=" style="width: 300px;" class="fr-fic fr-dib"><span class="fr-inner" spellcheck="false">Semiconductors must achieve at least 1,000x improvement in energy efficiency.</span></span></span></p><section id="isPasted"><h2>Optimizing Power for Billion-Plus Gate Designs</h2><p>An end-to-end approach to energy efficiency for AI accelerators must start at the architectural and micro-architectural levels during the earliest stages of the design flow and conclude at signoff. That’s why AI chip designers rely on architectural exploration platforms to map and evaluate power, performance, and area (PPA) tradeoffs for specific training or inference applications while proactively identifying critical vectors for downstream analysis.<br>As<strong>&nbsp;</strong>AI hardware typically consists of large arrays with thousands of tiles (processing elements), billion-plus-gate designs require multi-domain hardware and software power verification to minimize energy consumption and leakage. However, analyzing crucial power blocks and time windows requires advanced emulation systems to run billions of cycles and rapidly deliver multiple—and accurate—iterations. Only after completing this step can register transfer level (RTL) power analysis and physical implementation tools effectively optimize dynamic (gate switching) and static (leakage) power dissipation.<br><br>To consistently deliver accurate results, RTL power analysis tools for AI chip design should include the following capabilities:</p><ul><li>Timing-driven fast synthesis: Internal power calculation errors are often caused by fanout-based fast synthesis tools that fail to properly size cells following timing constraints. Like their downstream place-and-route counterparts, fast synthesis embedded in RTL power analysis tools must be timing driven.</li><li>Physically aware fast synthesis: RTL power analysis tools should be “physical aware” and capable of obtaining precise net capacitance values by executing first-pass placement of the cells in the design, as well as global routing. Unlike a fanout-based approach, physically aware capacitance estimation results in a unique and accurate value for each net.</li><li>Signoff-quality power computation engine: Traditional RTL power analysis tools using word-level logic inferencing for fast synthesis can only apply heuristic—and therefore inaccurate—methods for <a href="https://www.synopsys.com/ai/what-is-glitch-power.html#:~:text=Glitches%20occur%20if%20signal%20timing%20within%20the%20paths,leading%20to%20up%20to%2040%25%20additional%20dynamic%20power%29." target="_blank">glitch power computation</a>. To accurately calculate glitch power (which can potentially consume up to 40% of a chip’s total power) and reduce highly replicated tiles, RTL power analysis tools must have a signoff-quality power analysis engine, a netlist level design representation, and an integrated timing engine.</li></ul><p><br>After completing RTL power analysis and reduction, physical implementation (synthesis and place and route) tools can be used to further optimize PPA. To ensure reliability, scalability, and a frictionless user experience, these implementation tools should include a single, integrated data model architecture, interleaved engines, and a unified shell. Just as importantly, implementation tools should be capable of accurately modeling advanced node effects and glitch power to accelerate engineering change orders (ECOs) and final design closure.</p></section><section><h2>Exceeding Energy-Efficiency and Performance</h2><p>Synopsys offers a comprehensive end-to-end power solution to help AI chip designers cost-effectively meet or exceed ambitious performance and energy-efficiency goals while<a target="_blank"></a><a target="_blank">&nbsp;</a>accelerating time to market. Used at the very beginning of the design flow, <a href="https://www.synopsys.com/verification/virtual-prototyping/platform-architect.html" target="_blank">Synopsys Platform Architect™</a> provides AI chip designers with SystemC™ transaction-level modeling (TLM) tools and efficient methods to rapidly model, analyze, and optimize complex silicon architectures.&nbsp;<a href="https://www.synopsys.com/verification/emulation.html#zebuempower" target="_blank">Synopsys ZeBu® Empower</a>, a fast power profiler, is used for the next stage of the AI chip design process: analyzing and debugging energy consumption—based on hundreds of millions of cycles—for real software workloads.<br><br>Leading semiconductor companies have significantly reduced power draw with Synopsys ZeBu Empower, including&nbsp;<a href="https://sima.ai/" target="_blank">SiMa.ai</a>, a Silicon Valley-based AI chip startup that designs high-performance, low-energy AI chips for the intelligent edge. Specifically, the company realized a 2.5x frames per second (FPS) per watt improvement for its SiMa.ai™ Low Power MLSoC™. During a presentation at the SNUG Silicon Valley 2023 conference this spring, Sounil Biswas, director of silicon engineering at SiMa.ai, noted that subsequent silicon validation demonstrated excellent correlation between Synopsys ZeBu Empower data and board measurements.<br><br>To complement ZeBu Empower and enable RTL design for low power, we offer&nbsp;<a href="https://www.synopsys.com/implementation-and-signoff/signoff/primepower.html" target="_blank">Synopsys PrimePower RTL</a>, an RTL power analysis and reduction tool that consistently achieves accurate results (within +/- 15% of post-route implementation) by pairing timing-driven, physically aware synthesis capabilities with an integrated computation engine. Synopsys PrimePower RTL also provides step-by-step guidance to help AI chip designers further minimize glitching and reduce overall power consumption. &nbsp;<br><strong><br></strong>Additional PPA optimization is achieved with <a href="https://www.synopsys.com/implementation-and-signoff/physical-implementation/fusion-compiler.html" target="_blank">Synopsys Fusion Compiler™</a>, a comprehensive and integrated RTL-to-GDSII implementation system. After passing this milestone, the AI chip design is analyzed with <a href="https://www.synopsys.com/implementation-and-signoff/signoff/primepower.html" target="_blank">Synopsys PrimePower</a>, the golden power signoff solution. Certified by leading foundries worldwide down to 3nm processes, Synopsys PrimePower delivers fast runtime performance with distributed processing, achieving high accuracy at signoff within a few percent of SPICE and silicon measurements.</p></section><section><h2>Designing Differentiated Silicon for Edge AI Inference</h2><p>AI accelerators enable many popular applications to quickly analyze massive amounts of information and accurately infer results in milliseconds. At the same time, achieving optimal performance-per-watt remains a top priority for chip designers. This is especially true at the edge, where performance is often limited by minimal power envelopes and smaller die sizes.<br><br>However, these constraints create new opportunities for semiconductor companies to design differentiated silicon by precisely calibrating PPA to match the specific requirements of low-latency, high-bandwidth applications. For example, autonomous navigation demands a computational response latency limit of 20μs, while voice and video assistants must understand spoken keywords in less than 10μs and hand gestures in a few hundred milliseconds. To successfully implement PPA tradeoffs, chip designers should adopt a holistic approach to power optimization by leveraging an end-to-end solution that spans early architectural exploration to golden power signoff.<br><br>You can learn more about Synopsys <a href="https://wevlv.co/synopsys-energy-efficient" target="_blank" rel="noopener noreferrer">energy-efficient SoCs solutions here</a>.</p></section>

Leveraging Power Analysis for Energy-Efficient AI Accelerator

Designing Energy-Efficient AI Accelerators for Data Centers and the Intelligent Edge

Qualcomm's Pioneering Approach to Hybrid AI: Merging Cloud Processing with On-Device Efficiency and Scale


Balancing the Cloud and the Edge: A Close look at Enabling Hybrid AI

Sensors to monitor and regulate indoor air quality can have a positive impact on home, work, and school environments. Image Source: EPA 

Air Quality Indexes (AQIs) are used to measure the concentration of pollutants like particulate matter and other harmful substances in the air outside. 

Breathing Better: Using Environmental Sensors to Improve Indoor Air Quality

Single-board computers are an increasingly attractive option to shift electricity grid management towards more advanced and less resource consuming smart grids.


Enhancing Electrical Grid Management with emerging SBC technology

<p id="isPasted">If you&rsquo;re developing an IoT device, odds are that you want it to do some valuable computations to solve an important problem. Maybe you want to <a href="https://www.mistywest.com/posts/5-reasons-to-use-an-iot-framework/" target="_blank">deploy sensors in remote locations</a>, develop a device that can perform data analytics to monitor a renewable energy source, or build a medical device that can use computer vision to detect early signs of an illness.</p><p>Whatever you&rsquo;re building, at some point you may start to wonder: should your device perform these important computations in the cloud or at the edge? Choosing between computing on the cloud or on the edge is a decision that can impact things like your device&rsquo;s cost or efficiency &ndash; and no one wants to make the wrong decision initially and spend time and money later down the line to pivot to the correct one.&nbsp;</p><p>We asked some of our experienced engineers what they think someone needs to consider when choosing between computing in the cloud or on the edge.</p><p>&nbsp;</p><h2>What is cloud computing?</h2><p>The &ldquo;cloud&rdquo; refers to the collection of servers that can be accessed over the internet &ndash; popular <a href="https://www.mistywest.com/posts/azure-vs-aws-which-iot-framework-is-right-for-you/" target="_blank">cloud providers include Amazon Web Services, Microsoft Azure</a>, and Google Cloud.&nbsp;</p><p>These servers can provide on-demand computing resources to store and process data. You can think of the cloud as a centralized location for your files and programs, and you can connect any device to the cloud to access them. Services like Dropbox or Google Drive are some of the many cloud-based services out there.</p><p>&nbsp;</p><p><img data-fr-image-pasted="true" src="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-03.png" alt="" width="100%" height="auto" srcset="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-03.png 1360w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-03-300x155.png 300w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-03-1024x528.png 1024w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-03-768x396.png 768w" sizes="(max-width: 1360px) 100vw, 1360px" class="fr-fic fr-dii"></p><p>&nbsp;</p><p><em>Cloud computing</em> describes the idea of performing computations in the cloud. These computations can include data analysis and visualization, computer vision and machine learning. An example of cloud computing in action is when your average smart home speaker sends your audio input to the cloud where it is interpreted by algorithms and sends back a response. &nbsp;&nbsp;</p><p>&nbsp;</p><h2>What is edge computing?</h2><p><a href="https://blogs.nvidia.com/blog/2022/01/05/difference-between-cloud-and-edge-computing/#:~:text=The%20main%20difference%20between%20cloud,deploy%20them%20at%20the%20edge" target="_blank">The edge describes the &ldquo;edge&rdquo; of the network.&nbsp;</a>It includes devices that are an entry/exit point to the cloud, but are not part of the cloud itself. For example, a server in a data centre is part of the cloud; your smartphone and router that connects to that server is part of the edge.&nbsp;</p><p>&nbsp;</p><p><img data-fr-image-pasted="true" src="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-02.png" alt="" width="100%" height="auto" srcset="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-02.png 1360w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-02-300x155.png 300w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-02-1024x528.png 1024w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-02-768x396.png 768w" sizes="(max-width: 1360px) 100vw, 1360px" class="fr-fic fr-dii"></p><p>&nbsp;</p><p><em>Edge computing</em> describes the idea of performing computations on the edge. This way, the processing is done closer to or at the location where the data is collected or acted upon.</p><p>An example of an edge computing process is object detection on an autonomous vehicle. The vehicle processes data from its sensors and uses the results to avoid obstacles. Unlike your smart home speaker, the data it collects is processed locally, rather than sending it to the cloud.&nbsp;</p><p>&nbsp;</p><h1>Key Considerations</h1><p>There are a couple of key questions to consider when choosing between the edge and cloud computing.</p><p>&nbsp;</p><h2>What is the quality of your device&rsquo;s network?</h2><p>Performing computations on the cloud can work well when you have high-bandwidth, low-latency, and stable connection to the internet, as you will be needing to send your data back and forth between cloud servers and your device. If your device is intended to be used, for instance, in a home or office with a good internet connection, this back and forth can be done relatively seamlessly.&nbsp;</p><p>In most cases, if the computation is done on the edge, it won&rsquo;t be affected by poor or lost internet connection in a remote location; the processing can continue since it is not computed in the cloud. You wouldn&rsquo;t want your vehicle&rsquo;s object detection to stop working while on a long road trip; that&rsquo;s one of the reasons why autonomous vehicles frequently perform computations like object detection on the edge.&nbsp;</p><p>&nbsp;</p><h2>How <em>quick</em> and how <em>often</em> does your data need to be processed?</h2><p>Edge computing can be ideal in cases where your customer needs response times from your device to be faster than what can be achieved with a decent network connection, such as monitoring vital components of a system. The latency from travel-time between the device and the cloud can be reduced or eliminated completely. As a result, the data can be processed right away. If the data processing itself is quick, you could achieve real-time responses from your device.&nbsp;</p><p>Cloud computing is beneficial when device use is intermittent. Smart homes devices are a good example of this again, where running computations in the cloud lets you share the same computing resources between multiple customers. This reduces costs by avoiding the need to provision your device with upgraded hardware to run the data processing.&nbsp;</p><p>&nbsp;</p><p><img data-fr-image-pasted="true" src="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-05.png" alt="" width="100%" height="auto" srcset="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-05.png 1360w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-05-300x155.png 300w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-05-1024x528.png 1024w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-05-768x396.png 768w" sizes="(max-width: 1360px) 100vw, 1360px" class="fr-fic fr-dii"></p><p>&nbsp;</p><h2>What part of your data is important to you?</h2><p>Computing on the edge is useful if you only care about the result of your data <em>after</em> it has been processed. You can send only what is important to store in the long-term in the cloud, and doing so will allow you to reduce the cost of storing and processing data in the cloud. For example, if you are creating a traffic surveillance device that needs to report levels of congestion on a road, you could preprocess the videos on the edge&ndash;Instead of running hours of raw video in the cloud&ndash;and only send images or clips of the traffic when traffic is present.</p><p>It&rsquo;s possible you need to keep the data to <a href="https://www.mistywest.com/posts/improving-dataset-creation-for-machine-learning/" target="_blank">build out your machine learning dataset</a>, or you plan on analyzing the raw data in other ways in the future. If you are already sending your raw data to the cloud, it may be ideal to perform calculations in the cloud as well.</p><p>&nbsp;</p><h2>What are your device&rsquo;s power and size limitations?</h2><p>If you expect your device will be restricted in power and size, given that it has a good network connection, sending the computing work to be done on the cloud will allow your device to remain small and low-power. Google Home and Amazon Alexa, for example, will capture the audio and send it to the cloud for processing, allowing complex computations to be run on the audio that would not be possible to run on the small computers inside the devices themselves.</p><p>&nbsp;</p><p><img data-fr-image-pasted="true" src="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-04.png" alt="" width="100%" height="auto" srcset="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-04.png 1360w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-04-300x155.png 300w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-04-1024x528.png 1024w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest_Edge_Vs_Cloud-04-768x396.png 768w" sizes="(max-width: 1360px) 100vw, 1360px" class="fr-fic fr-dii"></p><p>&nbsp;</p><h2>Is your data processing model your Intellectual Property?</h2><p>If you are making a consumer device and the method you are using to process data is part of your Intellectual Property (IP), you may need to consider how you plan to protect it. Putting your IP on your device without a robust security plan can leave it vulnerable to hacks. If you don&rsquo;t have the knowledge or resources to secure your IP on the edge, it may be best to leave it on the cloud, which already has security measures in place.</p><p>&nbsp;</p><p><a data-rel="lightbox-image-0" data-rl_caption="" data-rl_title="" href="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest-Cloud-Vs-Edge-FlowChart-01.jpg" target="_blank" title=""><img data-fr-image-pasted="true" src="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest-Cloud-Vs-Edge-FlowChart-01.jpg" alt="" width="100%" height="auto" srcset="https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest-Cloud-Vs-Edge-FlowChart-01.jpg 1801w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest-Cloud-Vs-Edge-FlowChart-01-275x300.jpg 275w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest-Cloud-Vs-Edge-FlowChart-01-938x1024.jpg 938w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest-Cloud-Vs-Edge-FlowChart-01-768x839.jpg 768w, https://www.mistywest.com/wp-content/uploads/2022/05/MistyWest-Cloud-Vs-Edge-FlowChart-01-1406x1536.jpg 1406w" sizes="(max-width: 1801px) 100vw, 1801px" class="fr-fic fr-dii"></a></p><p>&nbsp;</p><h2>Some final considerations for choosing between edge and cloud computing</h2><p>There are quite a few things to consider when choosing between computing on the edge or in the cloud. In complex problems, you may benefit from using a combination of both, by leaving some parts of your processing on the edge and the rest on the cloud.</p><p>We created the above flow chart to help guide you through which path you should take, but there may always be more to consider for your device, depending on how complex your use case is.&nbsp;</p><p><br></p><p id="isPasted"><em>This article was originally written by Amanda Luu, former Product Design Engineer Coop at MistyWest,&nbsp;</em><a href="https://www.mistywest.com/posts/edge-vs-cloud-computing-which-solution-is-better-for-your-connected-device/" rel="noopener noreferrer" target="_blank"><em>for the MistyWest blog</em></a><em>.</em></p>

Choosing between computing on the cloud or on the edge is a decision that can impact things like your device’s cost or efficiency. Here’s how you can ensure you make the right decision early on in your product development roadmap.

Low power consumption

Latest Posts