News/Energy
DOE Launches $12B AI Initiative for Science, Energy, and Security Breakthroughs
The Department of Energy (DOE) has unveiled a $12 billion, five-year road map to develop integrated scientific AI systems, aiming to advance scientific discovery, energy solutions, and national security applications. Key focus areas of the FASST initiative: The Frontiers in Artificial Intelligence for Science, Security and Technology (FASST) project will establish a network of AI research clusters utilizing DOE's computing infrastructure, focusing on four main pillars: Data: Developing methods, platform protocols, and tools for efficient and safe aggregation and distribution of AI training datasets. Computing Infrastructure and Platforms: Developing and deploying next-generation AI computing platforms and infrastructure such as testbeds...
read Jul 19, 2024The AI Boom and Global Temperature Increases Are Fueling Unprecedented Electricity Usage
A surge in global electricity demand, fueled by rising temperatures, economic growth, and the AI boom, is set to reach its fastest pace in nearly two decades, according to a new report from the International Energy Agency. Record-breaking demand growth: The IEA forecasts that global electricity demand will grow by approximately 4% in 2024 and 2025, the highest annual growth rate since 2007, excluding exceptional rebound years following the global financial crisis and COVID-19 pandemic. Intense heatwaves in the first half of 2024 have already boosted demand and strained electricity systems, with increasing use of air conditioning expected to be...
read Jul 14, 2024Balancing Innovation and The Environmental Impact of AI
The rapid growth of artificial intelligence (AI) has raised concerns about its environmental impact, with the carbon footprint of AI becoming an increasingly pressing issue as the technology advances and becomes more widely adopted. Understanding AI's carbon footprint: To grasp the environmental implications of AI, it's important to consider the full lifecycle of AI systems, from hardware production to usage to deployment: Hardware production, maintenance, and recycling account for an estimated 30% of AI's carbon footprint, while computational costs make up the remaining 70%. Training large language models like GPT-3 can generate over 600,000 kg of CO2 equivalent (CO2e), comparable...
read Jul 8, 2024DeepMind Just Made a Breakthrough in AI Training
Google DeepMind's JEST AI training method promises significant speed and efficiency gains over traditional techniques, potentially addressing concerns about AI's growing power demands. Key Takeaways: DeepMind's JEST (joint example selection) training method breaks from traditional AI training by focusing on entire batches of data instead of individual data points: A smaller AI model first grades data quality from high-quality sources and ranks batches by quality. The small model then determines the batches most fit for training a larger model, resulting in up to 13 times faster training with 10 times less computation. Addressing AI's Power Demands: The JEST research comes...
read Jul 6, 2024Neuromorphic AI Chips Poised to Change Edge Computing and Intelligent Devices
The rise of neuromorphic computing and ultra-efficient AI chips promises to revolutionize edge computing and intelligent devices by dramatically reducing the energy and computational requirements of AI. Mimicking the brain's efficiency: Neuromorphic processors, like Innatera's Spiking Neural Processor, are designed to emulate the way biological brains process information using artificial neurons that communicate through spikes, enabling complex AI tasks with a fraction of the energy used by traditional solutions. These brain-inspired architectures are particularly well-suited for edge computing applications in consumer devices and industrial IoT, such as always-on audio processing, real-time sensor fusion, and ultra-low power computer vision. Innatera's neuromorphic...
read Jul 5, 2024AI Scalability Concerns: Experts Explore Solutions for Infrastructure Challenges and Future Potential
The rapid advancement of AI is creating concerns about the scalability of the infrastructure required to support it, potentially limiting its future potential. Key stakeholders are exploring various solutions to address the growing compute and storage demands while controlling costs and environmental impact. The challenge of scaling AI infrastructure: As large language models (LLMs) continue to grow, so do the training and inference requirements, raising concerns about the availability of GPU AI accelerators and the overall scalability of AI workloads: Daniel Newman, CEO at The Futurum Group, highlights the complexities that come with scaling AI, including the availability of power...
read Jul 5, 2024Softbank’s $10B AI Gambit: Betting Big on Chips, Energy, and the Future
Softbank, the Japanese conglomerate known for its aggressive bets on emerging technologies, is seeking to borrow $10 billion to fund AI-related projects, signaling a major push into the rapidly evolving AI landscape. Key focus areas: Softbank's AI investment strategy appears to be centered around two critical components of the AI ecosystem: Nvidia chips, which have become the de facto standard for AI training and inference, are a key target for Softbank's investments, as the company aims to capitalize on the growing demand for AI hardware. Energy startups are another area of focus, as the massive energy consumption of AI systems...
read Jul 4, 2024AI Hackathon Reveals Rapid Problem-Solving Potential, Heralds Specialized AI Infrastructure Era
The AI-powered hackathon hosted by Crusoe Energy and Lowercarbon Capital showcased the transformative potential of artificial intelligence in tackling complex challenges in the clean energy sector and beyond. AI accelerating problem-solving in the energy industry: The winning projects demonstrated how AI can compress months or years of traditional work into mere hours, potentially revolutionizing clean energy deployment: Verdigris developed an AI system that analyzes mortgage data to identify homeowners for zero-cost energy upgrades and generates personalized marketing materials, including AI-created images of homes with proposed improvements. Daylight created a system that rapidly extracts and maps complex stakeholder relationships from dense...
read Jul 4, 2024Google’s AI Surge Threatens Its 2030 Net-Zero Goal as Emissions Soar
Google's greenhouse gas emissions surge as AI investments grow, casting doubt on 2030 net-zero target. Emissions on the rise: Google's greenhouse gas emissions have increased significantly in recent years, driven largely by the company's investments in AI and associated infrastructure: Total emissions reached 14.3 million tonnes of carbon equivalent in 2023, a 48% increase from 2019 levels and a 13% rise since last year. Energy-related emissions, primarily from data center electricity consumption, rose 37% year-on-year and now represent a quarter of Google's total greenhouse gas emissions. Supply chain emissions, which account for 75% of the company's total emissions, also increased...
read Jul 3, 2024Crypto Miners Pivot to AI, Fueling Surge in Energy Demand
The soaring demand for AI computing is driving an energy gold rush, with crypto miners well-positioned to capitalize on their advanced equipment and access to low-cost power. Key Takeaways: The AI boom is causing a surge in demand for heavy-duty computing capacity and cheap energy: AI applications like ChatGPT require 10 times the electricity of traditional Google searches, making access to sufficient power vital for companies looking to meet the growing demand. Crypto miners, with their advanced equipment and access to low-cost energy in states like Texas and North Dakota, are becoming attractive partners for AI companies seeking to expand...
read Jul 3, 2024Google’s AI Emissions Surge Threatens Net-Zero Goal, Raises Sustainability Concerns
Google's AI-driven emissions increase raises concerns amid net-zero goals: Google's greenhouse gas emissions rose 13% last year, largely due to the increasing energy demands of artificial intelligence (AI), despite the company's aim to become carbon neutral by 2030. AI adoption poses challenges for emissions reduction efforts: The rapid growth and integration of AI into Google's products is driving up the company's energy consumption and associated emissions, making it more difficult to meet its sustainability targets: Google's data center electricity consumption grew by 17% last year, accounting for 7-10% of global data center electricity usage. Since 2019, Google's total emissions have...
read Jul 1, 2024AI and Genomics Breakthroughs Amid Geopolitical Tensions and Energy Challenges
Gloomy data centers, sunny solar power: The rapid growth of AI could strain US power grids, with data centers potentially consuming 8% of total electricity by 2030. However, this gloomy projection may be overstated, as renewable energy is rapidly expanding, with California already generating over 100% of its electricity from renewables at times. The data center boom and its energy demands will likely be a temporary issue, as energy mix, efficiency improvements, and evolving economics of both data centers and energy production are not fully accounted for in linear projections. Ultimately, the computational power of these data centers will be...
read Jun 27, 2024AI Breakthrough: Language Models without Matrix Multiplication, Slashing Power Consumption
Researchers claim a breakthrough in AI efficiency by eliminating matrix multiplication, a fundamental operation in current neural networks, which could significantly reduce the power consumption and costs of running large language models. Key Takeaways: Researchers from UC Santa Cruz, UC Davis, LuxiTech, and Soochow University have developed a method to run AI language models without using matrix multiplication (MatMul), which is currently accelerated by power-hungry GPU chips. Their custom 2.7 billion parameter model achieved similar performance to conventional large language models while consuming far less power when run on an FPGA chip. This development challenges the prevailing paradigm that matrix...
read Jun 26, 2024AI Data Center Boom: Powering Growth, Benefiting REITs, Energy, and Mining Sectors
The AI data center boom is set to drive significant growth in electricity demand and benefit a range of sectors, according to Jefferies analysts: Key takeaways: The increasing use of GPUs for AI investments has led to a surge in demand for data center space and power, with growth exceeding 30% in most markets over the past two years and showing no signs of slowing down. While Jefferies expects data center growth to remain robust, power generation constraints, supply chain issues, and labor market limitations may limit the acceleration rate. As the data center market expands, electricity demand is expected...
read Jun 26, 2024AI Breakthrough: MatMul-Free Language Modeling Slashes Energy Use, Boosts Accessibility
Researchers claim a breakthrough in AI efficiency by eliminating matrix multiplication, a fundamental operation in neural networks that is accelerated by GPUs, which could significantly reduce the energy consumption and costs of running large language models. Key innovation: MatMul-free language modeling; The researchers developed a custom 2.7 billion parameter model that performs similarly to conventional large language models (LLMs) without using matrix multiplication (MatMul): They demonstrated a 1.3 billion parameter model running at 23.8 tokens per second on a GPU accelerated by a custom FPGA chip, using only about 13 watts of power. This approach challenges the prevailing paradigm that...
read Jun 26, 2024AI’s Growing Energy Demand: A Nuanced Look Beyond the Hype
The energy demands of generative AI have recently come under scrutiny, but a closer look reveals a more nuanced picture: Data centers' growing energy use predates AI boom: While recent articles have focused on AI's impact on energy usage, the growth in data center energy consumption has been steady for over a decade, largely preceding the current generative AI boom: Worldwide data center energy usage has grown from about 100 TWh in 2012 to around 350 TWh in 2024, with the majority of this growth occurring before the launch of popular generative AI tools in 2022. Projections suggest that AI...
read Jun 26, 2024AI’s Energy Apocalypse Fears Exaggerated: Putting Generative AI’s Power Consumption Into Perspective
The rapid rise of generative AI has sparked concerns about its potential impact on global energy consumption and the power grid, but a closer examination suggests these fears may be overstated: Data centers' soaring energy use not solely due to AI: While recent media reports have focused on AI's energy demands, the growth in data center power usage largely predates the current generative AI boom. Data center energy usage has been growing steadily since 2012, with the majority of the increase occurring before the launch of popular AI tools like Dall-E and ChatGPT in 2022. The vast majority of data...
read Jun 21, 2024AI’s Power Surge Strains Grid, Spurs Fossil Fuel Use Despite Green Pledges
The rapid growth of artificial intelligence is straining the electricity grid, leading tech companies to pursue ambitious clean energy projects that may be overly optimistic, while fossil fuel use expands to meet the soaring demand. The AI power crunch: AI's voracious appetite for electricity is driving a nationwide data center building boom that is challenging the power grid and the tech industry's sustainability commitments: Training AI models and executing even simple AI tasks require increasingly complex computations that consume significantly more electricity compared to traditional computing. Major tech companies like Microsoft, Google, Amazon, and Meta are constructing sprawling data center...
read