AI Data Center Development
Plan, finance, design, construct, commission, and scale resilient high-density AI infrastructure from site strategy through operational handover.
AI Factories and the Development Lifecycle
Learning objective
By the end of this lesson, the learner will be able to explain ai factories and the development lifecycle, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
AI data centers unite accelerated compute, high-speed fabrics, storage, power, cooling, security, and software operations. Development proceeds through strategy, feasibility, design, procurement, construction, commissioning, and handover.
Instructor teaching guidance
Begin by establishing what the learner already knows about ai factories and the development lifecycle. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Build a stage-gate map for a proposed 20 MW AI facility and name the decision required at each gate.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
A successful project treats the facility, IT stack, grid, water, network, and operating model as one integrated system.
Knowledge check
Ask the learner to explain why “A successful project treats the facility, IT stack, grid, water, network, and operating model as one integrated system.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Demand, Workloads, and Capacity Models
Learning objective
By the end of this lesson, the learner will be able to explain demand, workloads, and capacity models, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Training, fine-tuning, inference, research, and mixed tenancy create different utilization, latency, redundancy, and density profiles. Capacity models should translate workload demand into racks, megawatts, storage, fabric bandwidth, and phased growth.
Instructor teaching guidance
Begin by establishing what the learner already knows about demand, workloads, and capacity models. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Create low, base, and high demand cases for a three-year AI capacity plan.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
Size infrastructure from workload evidence and uncertainty ranges, not from a single headline GPU count.
Knowledge check
Ask the learner to explain why “Size infrastructure from workload evidence and uncertainty ranges, not from a single headline GPU count.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Site Selection and Due Diligence
Learning objective
By the end of this lesson, the learner will be able to explain site selection and due diligence, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Evaluate utility capacity and delivery dates, fiber diversity, water and climate, land and geotechnical conditions, hazards, permitting, taxes, workforce, logistics, community impact, and expansion rights.
Instructor teaching guidance
Begin by establishing what the learner already knows about site selection and due diligence. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Score three hypothetical sites with a weighted due-diligence matrix.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
Cheap land cannot compensate for unavailable power, weak connectivity, or unmanageable development risk.
Knowledge check
Ask the learner to explain why “Cheap land cannot compensate for unavailable power, weak connectivity, or unmanageable development risk.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Power Architecture for Dynamic AI Loads
Learning objective
By the end of this lesson, the learner will be able to explain power architecture for dynamic ai loads, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Design utility service, substations, switchgear, UPS, generators or alternate energy, protection, grounding, power quality, and distribution for high-density loads with fast fluctuations.
Instructor teaching guidance
Begin by establishing what the learner already knows about power architecture for dynamic ai loads. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Sketch a one-line power path from grid interconnect to an AI rack and identify every resilience boundary.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
AI power design must address capacity, availability, quality, ramp behavior, harmonics, and maintainability together.
Knowledge check
Ask the learner to explain why “AI power design must address capacity, availability, quality, ramp behavior, harmonics, and maintainability together.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Cooling High-Density Compute
Learning objective
By the end of this lesson, the learner will be able to explain cooling high-density compute, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Compare air, rear-door heat exchangers, direct-to-chip liquid, and immersion approaches. Coordinate facility water systems, CDUs, heat-rejection equipment, water quality, leak detection, temperature classes, and heat reuse.
Instructor teaching guidance
Begin by establishing what the learner already knows about cooling high-density compute. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Choose a cooling topology for 25 kW, 60 kW, and 120 kW rack zones and defend each choice.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
Cooling selection follows rack density, hardware compatibility, climate, water strategy, serviceability, and lifecycle economics.
Knowledge check
Ask the learner to explain why “Cooling selection follows rack density, hardware compatibility, climate, water strategy, serviceability, and lifecycle economics.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
AI Compute, Fabric, Storage, and Data
Learning objective
By the end of this lesson, the learner will be able to explain ai compute, fabric, storage, and data, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Integrate accelerators, CPU hosts, scale-up and scale-out networking, data processing units, parallel storage, object storage, checkpoints, ingestion, and data-governance boundaries.
Instructor teaching guidance
Begin by establishing what the learner already knows about ai compute, fabric, storage, and data. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Draw the data path from source ingestion through training, checkpointing, validation, and inference.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
AI performance depends on the balance of compute, memory, network, storage, and software—not accelerators alone.
Knowledge check
Ask the learner to explain why “AI performance depends on the balance of compute, memory, network, storage, and software—not accelerators alone.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Resilience, Tier Intent, and Failure Domains
Learning objective
By the end of this lesson, the learner will be able to explain resilience, tier intent, and failure domains, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Translate business availability needs into redundancy, concurrent maintainability, compartmentation, fault domains, recovery objectives, spares, and tested operating procedures.
Instructor teaching guidance
Begin by establishing what the learner already knows about resilience, tier intent, and failure domains. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Perform a tabletop failure analysis for loss of one utility feed, one CDU, and one fabric switch.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
Resilience is an end-to-end service property; redundant components do not guarantee a resilient system.
Knowledge check
Ask the learner to explain why “Resilience is an end-to-end service property; redundant components do not guarantee a resilient system.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Sustainability, Energy, Water, and Carbon
Learning objective
By the end of this lesson, the learner will be able to explain sustainability, energy, water, and carbon, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Use PUE, WUE, CUE, renewable matching, grid carbon, heat reuse, embodied impacts, circularity, and workload scheduling to evaluate environmental performance without hiding tradeoffs.
Instructor teaching guidance
Begin by establishing what the learner already knows about sustainability, energy, water, and carbon. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Create an energy-and-water budget and identify three design levers with measurable targets.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
Efficient AI infrastructure optimizes useful computational work per total resource consumed.
Knowledge check
Ask the learner to explain why “Efficient AI infrastructure optimizes useful computational work per total resource consumed.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Cybersecurity and Physical Protection
Learning objective
By the end of this lesson, the learner will be able to explain cybersecurity and physical protection, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Apply zero trust, identity and access management, segmentation, secure management planes, supply-chain assurance, monitoring, incident response, cages, surveillance, access control, and environmental protection.
Instructor teaching guidance
Begin by establishing what the learner already knows about cybersecurity and physical protection. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Threat-model the path from a vendor laptop to a cluster management controller.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
Facilities technology, operational technology, IT, and AI platforms share risk and require coordinated controls.
Knowledge check
Ask the learner to explain why “Facilities technology, operational technology, IT, and AI platforms share risk and require coordinated controls.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Commercial Model, Finance, and Procurement
Learning objective
By the end of this lesson, the learner will be able to explain commercial model, finance, and procurement, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Model land, interconnection, construction, equipment, financing, energy, water, staffing, maintenance, refresh cycles, utilization, revenue, and residual value. Address long-lead gear and vendor concentration.
Instructor teaching guidance
Begin by establishing what the learner already knows about commercial model, finance, and procurement. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Build a risk-adjusted cost model with separate facility, IT, energy, and refresh assumptions.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
AI data center economics are dominated by schedule, utilization, power availability, capital intensity, and technology refresh.
Knowledge check
Ask the learner to explain why “AI data center economics are dominated by schedule, utilization, power availability, capital intensity, and technology refresh.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Construction, Quality, and Commissioning
Learning objective
By the end of this lesson, the learner will be able to explain construction, quality, and commissioning, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Use design reviews, digital coordination, factory tests, site acceptance, quality records, integrated systems testing, load banks, failure injection, documentation, and operator training.
Instructor teaching guidance
Begin by establishing what the learner already knows about construction, quality, and commissioning. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Write an integrated test script for loss of power while liquid-cooled compute is operating.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
Commission the service under realistic failure conditions, not merely individual equipment.
Knowledge check
Ask the learner to explain why “Commission the service under realistic failure conditions, not merely individual equipment.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.
Capstone: Development Investment Case
Learning objective
By the end of this lesson, the learner will be able to explain capstone: development investment case, apply it to a realistic situation, and defend the chosen action with evidence.
Core lesson
Synthesize market need, workload, site, utility, design basis, delivery schedule, risk register, capital plan, sustainability targets, operating model, and go/no-go gates.
Instructor teaching guidance
Begin by establishing what the learner already knows about capstone: development investment case. Teach the principle in plain language, define unfamiliar terms, then connect it to the course goal. Ask the learner to explain the idea back in their own words before moving to application.
Applied exercise
Prepare a board-ready investment memorandum for a phased AI data center campus.
Facilitation sequence
Guide the learner through the application in five steps: describe the present situation, identify the lesson principle, choose an action, name the evidence to review, and define how success will be measured. Use follow-up questions when an answer is vague or unsupported.
Mastery standard
A defensible development case connects technical architecture to finance, schedule, risk, community, and measurable customer demand.
Knowledge check
Ask the learner to explain why “A defensible development case connects technical architecture to finance, schedule, risk, community, and measurable customer demand.” is the correct standard, then identify one limitation, one risk, and one piece of evidence that could change the answer. Do not mark mastery from confidence alone; require clear reasoning and practical application.