You are previewing Energy Efficient Servers: Blueprints for Data Center Optimization.
O'Reilly logo
Energy Efficient Servers: Blueprints for Data Center Optimization

Book Description

Energy Efficient Servers: Blueprints for Data Center Optimization introduces engineers and IT professionals to the power management technologies and techniques used in energy efficient servers. The book includes a deep examination of different features used in processors, memory, interconnects, I/O devices, and other platform components. It outlines the power and performance impact of these features and the role firmware and software play in initialization and control. Using examples from cloud, HPC, and enterprise environments, the book demonstrates how various power management technologies are utilized across a range of server utilization. It teaches the reader how to monitor, analyze, and optimize their environment to best suit their needs. It shares optimization techniques used by data center administrators and system optimization experts at the world’s most advanced data centers.

Table of Contents

  1. Cover
  2. Title
  3. Copyright
  4. About ApressOpen
  5. Contents at a Glance
  6. Contents
  7. About the Authors
  8. About the Technical Reviewers
  9. Contributing Authors
  10. Acknowledgments
  11. Chapter 1: Why Data Center Efficiency Matters
    1. An Industry’s Call to Action
      1. Data Center Infrastructure Energy Use
      2. Energy Proportional Server Efficiency
      3. Regulatory Environment
    2. Measuring Energy Efficiency
      1. SPECPower
    3. High Performance Computing Efficiency
    4. Energy Efficiency and Cost
    5. Summary
  12. Chapter 2: CPU Power Management
    1. Server CPU Architecture/Design
      1. CPU Architecture Building Blocks
      2. CPU Design Building Blocks
      3. Intel Server Processors
    2. Introduction to Power
      1. CPU Power Breakdown
      2. Frequency, Voltage, and Temperature Interactions
    3. Power-Saving Techniques
      1. Turn It Off
      2. Turn It Down
      3. Power-Saving Strategies
    4. CPU Power and Performance States
      1. C-States
      2. P-States
      3. T-States
      4. S-States and G-States
      5. S0ix
      6. Running Average Power Limit (RAPL)
      7. CPU Thermal Management
      8. CPU Power Management Summary
    5. Summary
  13. Chapter 3: Memory and I/O Power Management
    1. System Memory
      1. Memory Architecture Basics
      2. Devices and Ranks
      3. Memory Error Correction (ECC)
      4. Memory Capacity
      5. Device Power Characteristics
      6. DDR3 vs. DDR4
      7. RDIMMs, UDIMMs, SODIMMs, and LRDIMMs
      8. Memory Channel Interleave and Imbalanced Memory Configurations
    2. Power and Performance States
      1. CKE Power Savings
      2. Self-Refresh
      3. Voltage/Frequency
    3. DDR Thermal Management
      1. Monitoring Temperature
      2. Memory Throttling
    4. CPU DDRIO
    5. Workload Behavior
    6. Memory Reliability Features
    7. CPU I/Os
      1. CPU Interconnect
      2. PCIe
    8. Summary
  14. Chapter 4: Platform Power Management
    1. Platform Overview
      1. Common Platform Components
      2. Integration
      3. Platform Manageability
    2. CPU Sockets
      1. Node Controllers
    3. Memory Risers and Memory Buffer Chips
    4. Server Chipsets
      1. PCH and Platform Power Management
      2. PCH Power Management
      3. PCIe in Chipsets
      4. PCH Thermal Management
    5. Networking
      1. Ambient Temperature, TDP, and Thermal Management
      2. Attached Media
      3. LAN Power Management Features
    6. USB
      1. Link Power States
      2. Link Frequency/Voltage
    7. Storage
      1. Storage Servers and Power Management
      2. HDDs and SDDs
      3. SATA and SAS Drive Power Management
      4. Frequency/Voltage
      5. NVMe Drive Power Management
    8. Power Delivery
      1. Overview of Power Delivery
      2. Power Converter Basics
      3. Power Conversion Losses
    9. Thermal Management
      1. System Considerations
      2. Component Thermal Management Features
      3. Platform Thermal Management
      4. Fan Speed Control and Design
    10. Summary
  15. Chapter 5: BIOS and Management Firmware
    1. BIOS Firmware
      1. Microcode Update
      2. Advanced Configuration and Power Interface
      3. Setup Utility
    2. Management Firmware
      1. Node ManagerCapabilities
      2. IPMI
      3. ACPI Power Metering Objects
    3. Summary
  16. Chapter 6: Operating Systems
    1. Operating Systems
      1. C-state Control
      2. C-state Policy
      3. P-state Control
      4. P-state Policy
      5. T-state Control
      6. Global Power Policy
      7. Process Scheduling
      8. Memory Management
      9. Device Drivers
    2. Virtualization
      1. Power State Control
      2. Consolidation
      3. VM Migration
    3. Comparison of Operating Environments
      1. Microsoft Windows Server (including Hyper-V)
      2. Linux Distributions (including KVM)
      3. VMWare ESX and ESXi
    4. Summary
  17. Chapter 7: Monitoring
    1. Hardware Monitoring
      1. Fixed Counters
      2. Core Performance Monitors
      3. Uncore Performance Monitors
      4. Status Snapshots
      5. Counter Access and CounterConstraints
      6. Events and Metrics
    2. Management Controller Monitoring
      1. ComponentPower Sensors
      2. Synthetic Sensors
      3. Sensors and Events
    3. Software Monitoring
      1. Utilization and Processor Time
      2. Processor Power State Requests
      3. Scheduler, Processes, and Threads
      4. Interrupts
      5. Memory
      6. I/O
    4. Tools
      1. Health Checks
      2. Hardware Monitoring Tools
      3. Operating System Monitoring Tools
    5. Summary
  18. Chapter 8: Characterization and Optimization
    1. Workloads
      1. Identifying Suitable Workloads
      2. Workload Types
    2. System Characterization
      1. Steady State vs. Non-Steady State
      2. Data Collection
      3. Methodology
      4. Analysis
    3. Optimization
      1. CPU Power Management
      2. Memory
      3. NIC
      4. Storage
      5. Thermal Management
      6. Optimization at a Glance
    4. Summary
  19. Chapter 9: Data Center Management
    1. Data Center Management and Power Distribution
    2. Data Center Facilities
      1. Power Infrastructure
      2. Cooling Infrastructure
    3. Simplified Total Cost Models of Cost and Compute Infrastructure
    4. Performance per Watt per Dollar
    5. Summary
  20. Appendix A: Technology and Terms
  21. Index