AI Ecosystem CourseLearn with Q
AI Ecosystem · In Development

Data-Center Physical Infrastructure

See how data centres keep AI computers powered, connected and cool.

About This Course

Follow two paths: power in and heat out

A room of accelerators requires several supporting systems. What keeps them working reliably? The building is one part of the answer. Electrical equipment, cooling, controls, networks, maintenance, and operating procedures all have a role. A dependable service also needs tested software, networking, monitoring and operating procedures.

We will follow two paths: electricity from the campus connection to a rack, and heat from a processor to the surrounding environment. Then we will examine redundancy, network routes, operating readiness, and capacity claims. Revisit Energy And Power Generation for power and energy, and Compute Fundamentals for workload requirements.

Your project is a 100 MW AI Data Center Architecture. Define what the capacity measure includes, allocate IT capacity, trace cooling, identify where failures can spread, and specify how a customer would accept the service. This is a learning exercise. Construction requires professional engineering, equipment specifications, codes, and site data.

Planned Lessons

  1. Facility Load Is Larger Than Accelerator LoadSee why an AI site needs electricity for cooling and other equipment as well as its chips.
  2. The Substation Is Part Of The Delivery PathFollow electricity through the equipment that connects a data centre to the power network.
  3. Switchgear Controls And Protects CircuitsSee how switches and protective equipment help stop electrical faults from spreading.
  4. UPS And Backup Generation Serve Different Time HorizonsCompare backup equipment that responds immediately with generators used for longer outages.
  5. Distribution Reaches The RackFollow the power supply from the building’s equipment to the racks of AI computers.
  6. Computers Produce Heat That Must Move SomewhereSee where the heat from AI computers goes and why it must be removed.
  7. Air Cooling Depends On Controlled AirflowExplore how moving air through the right paths keeps servers cool.
  8. Direct-To-Chip Liquid Cooling And Cold PlatesSee how liquid can carry heat directly away from an AI chip.
  9. CDUs And Heat Exchangers Connect LoopsFollow heat as it moves between the cooling systems inside servers and the building.
  10. Chillers, Cooling Towers, And Dry Coolers Reject HeatCompare the equipment a data centre uses to release heat into the surrounding environment.
  11. Rack Density Changes The Building ProblemFind out why putting more computing power in one rack changes a building’s design.
  12. N, N+1, And 2N Need A Stated BoundaryUnderstand what extra backup equipment is included in different reliability claims.
  13. Failure Domains Connect Physical And Software DesignIdentify which parts of a data centre might stop working together after one failure.
  14. Fiber Routes, Geography, And LatencySee how distance and cable routes affect the speed and reliability of an AI service.
  15. Construction, Energization, And CommissioningFollow a data centre from construction through connection to power and final testing.
  16. Customer-Ready Capacity And Phase AcceptanceFind out what must be ready before a customer can use new computing capacity.
  17. Operations And Retirement Are Part Of The LifecycleInclude maintenance, replacement and closure when planning the life of a data centre.