Eaton – Eaton completes acquisition of leading liquid-cooling solutions provider Boyd Thermal, creating an industry-leading grid-to-chip solution for data centers

EMR Analysis

More information on Eaton: See full profile on EMR Executive Services

More information on Paulo Ruiz Sternadt (Chief Executive Officer, Eaton): See the full profile on EMR Executive Services

More information on Olivier Leonetti (Senior Leadership Team – Executive Vice President and Chief Financial Officer, Eaton till March 13, 2026): See full profile on EMR Executive Services

More information on David Foster (Senior Leadership Team – Executive Vice President and Chief Financial Officer, Eaton as from March 2, 2026): See full profile on EMR Executive Services

More information on Eaton’s 2030 Growth Strategy (Lead, Invest and Execute for Growth) by Eaton: See full profile on EMR Executive Services

 

More information on Boyd Thermal Business by Eaton:  https://www.boydcorp.com/thermal.html + https://www.eaton.com/us/en-us/products/thermal-management-solutions/eaton-and-boyd-thermal.html 

Boyd Thermal Management: Boyd has a long history of developing, designing, testing, optimizing, and fabricating reliable high-performance thermal management solutions across all industries. Through consistent innovation in engineering and manufacturing, Boyd provides optimized, cost-efficient cooling solutions and systems utilizing the largest range of traditional and advanced cooling technologies.

Boyd Thermal has forecasted sales of $1.7 billion for 2026, of which $1.5 billion is in liquid cooling.

Boyd Thermal is a global business based in the U.S., with more than 6,000 employees and manufacturing sites across North America, Asia and Europe. With its decades-long history starting as an aerospace thermal management supplier, today Boyd’s thermal business serves data center, industrial, aerospace and other markets. 

More information on David Huang (Chief Executive Officer, Boyd Thermal Business, Eaton): https://www.boydcorp.com/about-boyd/leadership.html + https://www.linkedin.com/in/david-huang-31960b7/ 

 

 

 

More information on Boyd Corporation: https://www.boydcorp.com/ +Boyd is the world’s leading innovator in sustainable engineered material and thermal solutions that make our customers’ products better, safer, faster, and more reliable. We develop and combine technologies to solve ambitious performance targets in our customers’ most critical applications. By implementing technologies and material science in novel ways to seal, protect, cool, and interface, Boyd has continually redefined the possible and championed customer success for over 90 years.

More information on Doug Britt (Chief Executive Officer, Boyd Corporation): https://www.boydcorp.com/about-boyd/leadership.html + https://www.linkedin.com/in/doug-britt-3b361b5/ 

 

 

 

More information on Goldman Sachs & Co. LLC.: https://www.goldmansachs.com/ + The Goldman Sachs Group, Inc. is a leading global investment banking, securities and investment management firm that provides a wide range of financial services to a substantial and diversified client base that includes corporations, financial institutions, governments and individuals. Founded in 1869, the firm is headquartered in New York and maintains offices in all major financial centers around the world.

More information on David Solomon (Chairman and Chief executive Officer, Goldman Sachs Inc.): https://www.goldmansachs.com/our-firm/our-people-and-leadership/leadership + https://www.linkedin.com/in/david-m-solomon/ 

More information on Goldman Sachs Asset Management by Goldman Sachs & Co. LLC.: https://am.gs.com/en-int/institutions + Bringing together traditional and alternative investments, Goldman Sachs Asset Management provides clients around the world with a dedicated partnership and focus on long-term performance.

As the primary investing area within Goldman Sachs, we deliver investment and advisory services for the world’s leading institutions, financial advisors and individuals, drawing from our deeply connected global network and tailored expert insights, across every region and market.

US$ 3.3+ Trillion Assets under supervision

1.7K+ investment professionals

More information on Marc Nachmann (Global Head of Asset & Wealth Management, Goldman Sachs Asset Management, Goldman Sachs Inc.): https://www.goldmansachs.com/our-firm/our-people-and-leadership/leadership#management-committee 

 

 

 

 

 

 

 

EMR Additional Notes:

• Grid, Microgrids, DERs and DERM’s:
  • Grid / Power Grid:
    • The power grid is a network for delivering electricity to consumers. The power grid includes generator stations, transmission lines and towers, and individual consumer distribution lines.
      • The grid constantly balances the supply and demand for the energy that powers everything from industry to household appliances.
      • Electric grids perform three major functions: power generation, transmission, and distribution.
  • Microgrid:
    • Small-scale power grid that can operate independently or collaboratively with other small power grids. The practice of using microgrids is known as distributed, dispersed, decentralized, district or embedded energy production.
  • Smart Grid:
    • Any electrical grid + IT at all levels.
  • Micro Grid:
    • Group of interconnected loads and DERs (Distributed Energy Resources) within a clearly defined electrical and geographical boundaries witch acts as a single controllable entity with respect to the main grid.
  • Distributed Energy Resources (DERs): 
    • Small-scale electricity supply (typically in the range of 3 kW to 50 MW) or demand resources that are interconnected to the electric grid. They are power generation resources and are usually located close to load centers, and can be used individually or in aggregate to provide value to the grid.
      • Common examples of DERs include rooftop solar PV units, natural gas turbines, microturbines, wind turbines, biomass generators, fuel cells, tri-generation units, battery storage, electric vehicles (EV) and EV chargers, and demand response applications.
  • Distributed Energy Resources Management Systems (DERMS):
    • Platforms which helps mostly distribution system operators (DSO) manage their grids that are mainly based on distributed energy resources (DER).
      • DERMS are used by utilities and other energy companies to aggregate a large energy load for participation in the demand response market. DERMS can be defined in many ways, depending on the use case and underlying energy asset.

 

 

• Chip, Computer Chip and Integrated Circuit (IC):
  • Assembly of electronic components in which hundreds to millions of transistors, resistors, and capacitors are interconnected and built up on a thin substrate of semiconductor material (usually silicon) to form a small chip or wafer.
  • Terms are often used interchangeably but there are subtle differences:
    • Chip: Is the most general term. It simply refers to a small piece of semiconductor material.
    • Computer Chip: This term is more specific. It refers to a chip designed for use in computers, such as microprocessors, memory chips, and graphics processing units (GPUs).
    • Integrated Circuit (IC): This is the technical term for the complex circuitry etched onto the semiconductor material. It describes the design and functionality of the electronic components on the chip.
• AI Chips:
  • Artificial intelligence (AI) chips are specially designed computer microchips used in the development of AI systems. Unlike other kinds of chips, AI chips are often built specifically to handle AI tasks, such as machine learning (ML), data analysis and natural language processing (NLP).
  • Chips are made of silicon, a semiconductor material.
  • According to The Economist, chipmakers on the island of Taiwan produce over 60% of the world’s semiconductors and more than 90% of its most advanced chips. Unfortunately, critical shortages and a fragile geopolitical situation are constraining growth.
    Nvidia, the world’s largest AI hardware and software company, relies almost exclusively on Taiwan Semiconductor Manufacturing Corporation (TSMC) for its most advanced AI chips.
• Grid-to-Chip:
  • “Grid-to-chip” is an expression used mainly in data-center and semiconductor power-electronics contexts.
  • It describes the entire end-to-end chain of electrical power delivery; the full power path from the electrical grid all the way to the silicon chip.

 

 

• Liquid Cooling Components:
  • Direct-to-Chip (DTC) Cooling:
    • A Direct-to-Chip (DTC) cooling system is a liquid-cooling technology used to cool high-performance computer chips—such as CPUs, GPUs, and accelerators—by bringing a liquid coolant directly to the chip surface through a cold plate.
    • It is one of the most efficient cooling methods used in data centers, HPC clusters, and AI server farms.
  • Chillers:  
    • Mechanical systems that remove heat from a building’s liquid coolant, typically water, and transfer it to another location to cool the air and maintain comfort. Unlike traditional systems that might cool air directly, chillers generate chilled water that circulates through air handling units (AHUs) within the space to absorb heat, making them essential for cooling large commercial or industrial buildings.
  • Technology Cooling System (TCS): 
    • Broad, non-standard term that generally refers to any integrated system, including chillers, designed to manage heat in technological applications, like data centers or industrial processes, to prevent overheating
  • Condensors:
    • A condenser is a heat exchanger that cools a gas or vapor, causing it to condense into a liquid, a process that releases latent heat. These devices are critical in many systems, such as air conditioning, refrigeration, and power plants, where they use air or water to absorb heat from the vapor, transforming it back into a liquid state.
  • CDU (Coolant Distribution Unit): 
    • A coolant distribution unit contains a pump that circulates coolant through a network of pipes or channels, distributing it to various components like servers, processors, or other high heat components in large, high power devices that need cooling.
    • Coolant Distribution Units are essential in data centers with high-density applications providing close controlled coolant delivery and precise control of the liquid cooling system. They help manage heat loads, reduce power consumption, and increase efficiency and reliability through redundancy.
  • HDU (Heat Dissipation Unit):
    • Unlike a Coolant Distribution Unit (CDU) that rejects heat from the server rack to the chilled water loop of the building, a Heat Dissipation Unit (HDU™) rejects the server heat from the server rack to the white space where building air conditioning would then remove that heat.
  • RDHX (Rear Door Heat Exchanger) – Rack-level Heat Exchange: 
    • Chilled water cooling doors that fit onto a IT rack. The air is pulled through the rack by the fans or pushed through by the server fans for the passive rear door (no fans).

 

 

• AI – Artificial Intelligence:
  • Artificial intelligence is the simulation of human intelligence processes by machines, especially computer systems.
  • As the hype around AI has accelerated, vendors have been scrambling to promote how their products and services use AI. Often what they refer to as AI is simply one component of AI, such as machine learning. AI requires a foundation of specialized hardware and software for writing and training machine learning algorithms. No one programming language is synonymous with AI, but several, including Python, R and Java, are popular.
  • In general, AI systems work by ingesting large amounts of labeled training data, analyzing the data for correlations and patterns, and using these patterns to make predictions about future states. In this way, a chatbot that is fed examples of text chats can learn to produce lifelike exchanges with people, or an image recognition tool can learn to identify and describe objects in images by reviewing millions of examples.
  • AI programming focuses on three cognitive skills: learning, reasoning and self-correction.
  • The 4 types of artificial intelligence?
    • Type 1: Reactive machines. These AI systems have no memory and are task specific. An example is Deep Blue, the IBM chess program that beat Garry Kasparov in the 1990s. Deep Blue can identify pieces on the chessboard and make predictions, but because it has no memory, it cannot use past experiences to inform future ones.
    • Type 2: Limited memory. These AI systems have memory, so they can use past experiences to inform future decisions. Some of the decision-making functions in self-driving cars are designed this way.
    • Type 3: Theory of mind. Theory of mind is a psychology term. When applied to AI, it means that the system would have the social intelligence to understand emotions. This type of AI will be able to infer human intentions and predict behavior, a necessary skill for AI systems to become integral members of human teams.
    • Type 4: Self-awareness. In this category, AI systems have a sense of self, which gives them consciousness. Machines with self-awareness understand their own current state. This type of AI does not yet exist.
  • Machine Learning (ML):
    • Developed to mimic human intelligence, it lets the machines learn independently by ingesting vast amounts of data, statistics formulas and detecting patterns.
    • ML allows software applications to become more accurate at predicting outcomes without being explicitly programmed to do so.
    • ML algorithms use historical data as input to predict new output values.
    • Recommendation engines are a common use case for ML. Other uses include fraud detection, spam filtering, business process automation (BPA) and predictive maintenance.
    • Classical ML is often categorized by how an algorithm learns to become more accurate in its predictions. There are four basic approaches: supervised learning, unsupervised learning, semi-supervised learning and reinforcement learning.
  • Deep Learning (DL):
    • Subset of machine learning, Deep Learning enabled much smarter results than were originally possible with ML. Face recognition is a good example.
    • DL makes use of layers of information processing, each gradually learning more and more complex representations of data. The early layers may learn about colors, the next ones about shapes, the following about combinations of those shapes, and finally actual objects. DL demonstrated a breakthrough in object recognition.
    • DL is currently the most sophisticated AI architecture we have developed.
  • Generative AI (GenAI):
    • Generative AI technology generates outputs based on some kind of input – often a prompt supplied by a person. Some GenAI tools work in one medium, such as turning text inputs into text outputs, for example. With the public release of ChatGPT in late November 2022, the world at large was introduced to an AI app capable of creating text that sounded more authentic and less artificial than any previous generation of computer-crafted text.
  • Small Language Models (SLM) and Large Language Models (LLM):
    • Small Language Models (SLMs) are artificial intelligence (AI) models capable of processing, understanding and generating natural language content. As their name implies, SLMs are smaller in scale and scope than large language models (LLMs).
    • LLM means Large Language Models — a type of machine learning/deep learning model that can perform a variety of natural language processing (NLP) and analysis tasks, including translating, classifying, and generating text; answering questions in a conversational manner; and identifying data patterns.
    • For example, virtual assistants like Siri, Alexa, or Google Assistant use LLMs to process natural language queries and provide useful information or execute tasks such as setting reminders or controlling smart home devices.
  • Computer Vision (CV) / Vision AI & Machine Vision (MV):
    • Field of AI that enables computers to interpret and act on visual data (images, videos). It works by using deep learning models trained on large datasets to recognize patterns, objects, and context.
    • The most well-known case of this today is Google’s Translate, which can take an image of anything — from menus to signboards — and convert it into text that the program then translates into the user’s native language.
    • Machine Vision (MV) :
      • Specific application for industrial settings, relying on cameras to analyze tasks in manufacturing, quality control, and worker safety. The key difference is that CV is a broader field for extracting information from various visual inputs, while MV is more focused on specific industrial tasks.
      • Machine Vision is the ability of a computer to see; it employs one or more video cameras, analog-to-digital conversion and digital signal processing. The resulting data goes to a computer or robot controller. Machine Vision is similar in complexity to Voice Recognition.
      • MV uses the latest AI technologies to give industrial equipment the ability to see and analyze tasks in smart manufacturing, quality control, and worker safety.
  • Multimodal Intelligence and Agents:
    • Subset of artificial intelligence that integrates information from various modalities, such as text, images, audio, and video, to build more accurate and comprehensive AI models.
    • Multimodal capabilities allows AI to interact with users in a more natural and intuitive way. It can see, hear and speak, which means that users can provide input and receive responses in a variety of ways.
    • An AI agent is a computational entity designed to act independently. It performs specific tasks autonomously by making decisions based on its environment, inputs, and a predefined goal. What separates an AI agent from an AI model is the ability to act. There are many different kinds of agents such as reactive agents and proactive agents. Agents can also act in fixed and dynamic environments. Additionally, more sophisticated applications of agents involve utilizing agents to handle data in various formats, known as multimodal agents and deploying multiple agents to tackle complex problems.
  • Agentic AI:
    • Agentic AI is an artificial intelligence system that can accomplish a specific goal with limited supervision. It consists of AI agents—machine learning models that mimic human decision-making to solve problems in real time. In a multiagent system, each agent performs a specific subtask required to reach the goal and their efforts are coordinated through AI orchestration.
    • Unlike traditional AI models, which operate within predefined constraints and require human intervention, agentic AI exhibits autonomy, goal-driven behavior and adaptability. The term “agentic” refers to these models’ agency, or, their capacity to act independently and purposefully.
    • Agentic AI builds on generative AI (gen AI) techniques by using large language models (LLMs) to function in dynamic environments. While generative models focus on creating content based on learned patterns, agentic AI extends this capability by applying generative outputs toward specific goals.
  • Edge AI Technology:
    • Edge artificial intelligence refers to the deployment of AI algorithms and AI models directly on local edge devices such as sensors or Internet of Things (IoT) devices, which enables real-time data processing and analysis without constant reliance on cloud infrastructure.
    • Simply stated, edge AI, or “AI on the edge“, refers to the combination of edge computing and artificial intelligence to execute machine learning tasks directly on interconnected edge devices. Edge computing allows for data to be stored close to the device location, and AI algorithms enable the data to be processed right on the network edge, with or without an internet connection. This facilitates the processing of data within milliseconds, providing real-time feedback.
    • Self-driving cars, wearable devices, security cameras, and smart home appliances are among the technologies that leverage edge AI capabilities to promptly deliver users with real-time information when it is most essential.
  • High-Density AI: 
    • High-density AI refers to the concentration of AI computing power and storage within a compact physical space, often found in specialized data centers. This approach allows for increased computational capacity, faster training times, and the ability to handle complex simulations that would be impossible with traditional infrastructure.
  • Explainable AI (XAI) and Human-Centered Explainable AI (HCXAI): 
    • Explainable AI (XAI) refers to methods for making AI model decisions understandable to humans, focusing on how the AI works, whereas Human-Centered Explainable AI (HCXAI) goes further by contextualizing those explanations to a user’s specific task and understanding needs. While XAI aims for technical transparency of the model, HCXAI emphasizes the human context, emphasizing user relevance, and the broader implications of explanations, including fairness, trust, and ethical considerations.
  • Physical AI & Embodied AI: 
    • Physical AI refers to a branch of artificial intelligence that enables machines to perceive, understand, and interact with the physical world by directly processing data from a variety of sensors and actuators.
    • Physical AI provides the overarching framework for creating autonomous systems that act intelligently in real-world settings. Embodied AI, as a subset, focuses on the sensory, decision-making, and interaction capabilities that enable these systems to function effectively in dynamic and unpredictable environments.
  • Federated Learning and Reinforcement Learning:
    • Federated Learning is a machine-learning technique where data stays where it is, and only the learned model updates are shared. “Training AI without sharing your data”.
    • Reinforcement Learning is a type of AI where an agent learns by interacting with an environment and receiving rewards or penalties. “Learning by trial and error”
    • Federated Learning (FL) and Reinforcement Learning (RL) can be combined into a field called Federated Reinforcement Learning (FRL), where multiple agents learn collaboratively without sharing their raw data. In this approach, each agent trains its own RL policy locally and shares model updates, like parameters or gradients, with a central server. The server aggregates these updates to create a more robust, global model. FRL is used in applications like optimizing resource management in communication networks and enhancing the performance of autonomous systems by learning from diverse, distributed experiences while protecting privacy.