How New Computing Technology Can Keep the Power On

The latest innovations in chip design, AI, and cybersecurity can help prevent grid failures and improve resilience

The recent winter storms that swept across the US exposed the vulnerability of the power grid, especially in Texas, where millions of people suffered from prolonged blackouts and soaring electricity bills. The Electric Reliability Council of Texas (ERCOT), which manages the grid for most of the state, faced criticism for its lack of preparedness and communication during the crisis.

While extreme weather events pose a serious threat to the grid, they are not the only challenge. The grid also has to cope with the increasing demand for electricity, the integration of renewable energy sources, the aging of infrastructure, and the risk of cyberattacks. To address these issues, new computing technology can play a vital role in enhancing the reliability and resilience of the grid.

Here are some of the new computing technologies that can help keep the power on:

• High-performance, energy-efficient chips. The development of artificial intelligence (AI) and large-scale data analysis requires a lot of computing power, which in turn consumes a lot of energy. To reduce the energy footprint of these applications, chip makers are innovating new ways to design and manufacture chips that are faster, smaller, and more efficient. For example, Intel has announced its plans to introduce new chip technologies, such as RibbonFET and PowerVia, that can improve the performance and functionality of transistors and power delivery12. These technologies can enable more advanced AI and data analytics for the grid, such as demand forecasting, load balancing, fault detection, and grid optimization.

• Deep learning with light. Another way to boost the computing power and efficiency of AI is to use light instead of electricity. Light-based computing, also known as photonic computing, can process information faster and with less heat than electronic computing. Researchers at MIT have developed a new computing architecture that uses light to perform deep learning computations on a low-power, memory-constrained device3. This technique can enable self-driving cars to make decisions in real-time while using a fraction of the energy currently demanded by their onboard computers. Similarly, this technique can enable smart grid devices, such as sensors, meters, and controllers, to perform complex tasks locally and autonomously, without relying on centralized servers or cloud services.

• Cybersecurity strategy for ERCOT. The power grid is a critical infrastructure that is constantly under the threat of cyberattacks, which can disrupt the supply and delivery of electricity, compromise the safety and privacy of customers, and cause economic and social damage. To protect the grid from cyberattacks, ERCOT and other grid operators need to adopt a comprehensive and proactive cybersecurity strategy that covers the grid's physical, operational, and informational aspects. A report by the University of Texas at Austin proposes a cybersecurity strategy for ERCOT that includes the following elements:

• Establishing a clear governance structure and accountability for cybersecurity within ERCOT and among its stakeholders.

• Implementing best practices and standards for cybersecurity, such as encryption, authentication, monitoring, and auditing.

• Enhancing the cybersecurity awareness and training of ERCOT staff and market participants.

• Collaborating with other grid operators, regulators, law enforcement, and research institutions to share information and resources on cybersecurity.

These are some of the new computing technologies that can help keep the power on and improve the resilience of the grid. By adopting and applying these technologies, ERCOT and other grid operators can better prepare for and respond to the challenges and opportunities that lie ahead.