ASHRAE Book ITEquipmentDesign 2016
$31.96
ASHRAE Datacom Series Book 13: IT Equipment Design Impact on Data Center Solutions
| Published By | Publication Date | Number of Pages |
| ASHRAE | 2016 | 142 |
With everything from smart phones to thermostats generating data, back-end IT systems are experiencing massive hardware demands. Data centers must have a footprint that is flexible, scalable, and adaptable. They must be able to move as fast as new applications are developed and keep up with new ideas, new architectures, and new ways of thinking—all in real time. This book equips facility planners, operators, IT equipment (ITE) manufacturers, HVAC&R manufacturers, and end users with the knowledge they need to select the equipment and design best suited to the modern and evolving data center. This book is the thirteenth in the ASHRAE Datacom Series, authored by ASHRAE Technical Committee 9.9, Mission Critical Facilities, Data Centers, Technology Spaces and Electronic Equipment. The series provides comprehensive treatment of datacom cooling and related subjects.
PDF Catalog
| PDF Pages | PDF Title |
|---|---|
| 17 | Figure 2.1 Processor package ingredients. Figure 2.2 Processor package with substrate and lid. |
| 18 | Figure 2.3 Processor heat sink for air-cooling. Figure 2.4 Processor heat sink for liquid-cooling (disassembled). |
| 19 | Figure 2.5 Typical dual in-line memory module (DIMM) with and without a heat spreader. Figure 2.6 (a) HDD with cover removed, (b) NVM small form factor SSD, and (c) a PCIe NVM SSD. |
| 20 | Figure 2.7 Dual processor socket motherboard. 2.1.1 Rack-Mount Servers |
| 21 | Figure 2.8 Typical compute server rack and packaging. 2.1.2 Blade Servers |
| 22 | Figure 2.9 1U dual-socket volume server Figure 2.10 Example of a blade. |
| 23 | Figure 2.11 Example blade server chassis. Figure 2.12 TOR switch with top cover removed. |
| 25 | Table 2.1 Common Types of Networking Equipment and Their Functions |
| 26 | Figure 2.13 Example: Network implementation in a small data center. Figure 2.14 Example: Network implementation in a medium-to-large data center. |
| 27 | Figure 2.15 Example: Network implementation with TOR switches. |
| 29 | Figure 2.16 Front and rear views of storage array enclosure. |
| 30 | Table 2.2 Summary of the Four Primary Functions of Information Storage |
| 31 | Figure 2.17 Schematic diagram of typical disk storage array. |
| 32 | Figure 2.18 Racks populated with 4U disk storage arrays. |
| 33 | (a) (b) Figure 2.19 Example of a single tape drive (left) and a tape media cartridge (right). |
| 34 | Figure 2.20 ASHRAE definitions for rack airflow. Source: Thermal Guidelines for Data Processing Environments (ASHRAE 2015d). Figure 2.21 Large switch with front-to-rear airflow. |
| 35 | Figure 2.22 Midsize switch with side-to-side airflow. Figure 2.23 Small networking equipment with “S” shaped airflow. |
| 36 | Figure 2.24 1 or 2U switch with side-to-side airflow. Figure 2.25 1 or 2U switch with front-to-rear airflow. |
| 37 | Figure 2.26 Small form factor equipment cooled with natural convection. |
| 40 | Figure 3.1 Evolution of processor module level heat flux in high-end servers. |
| 42 | Figure 3.2 System thermal management. |
| 43 | Figure 3.3 Component temperature effects. |
| 44 | 3.2.1 Component Temperature Limits 3.2.1.1 Reliability 3.2.1.2 Functional |
| 45 | 3.2.1.3 Damage 3.2.2 Key Individual Component Specifications 3.2.2.1 Processors 3.2.2.2 Memory |
| 46 | 3.2.2.3 Support Logic (Application-Specific Integrated Circuits [ASICs]) 3.2.2.4 Hard Disk Drives (HDDs) and Nonvolatile Memory (NVM) 3.3.1 Processors |
| 47 | Figure 3.4 System and processor states. 3.3.1.1 Processor States 3.3.1.2 Performance States |
| 48 | 3.3.1.3 Thermal States |
| 49 | Figure 3.5 Processor thermal control. 3.3.2 Memory |
| 50 | Figure 3.6 Memory thermal management. 3.3.3 Support Logic 3.3.4 Voltage Regulators (VRs) 3.3.5 Storage Devices |
| 51 | 3.4.1 System Design and Optimization 3.4.1.1 The Thermal Equation |
| 52 | 3.4.1.2 Fan Selection and Design Considerations |
| 53 | 3.4.1.3 Fan Laws |
| 54 | 3.4.1.4 Board and Chassis Layout Figure 3.7 1U systems. |
| 55 | 3.4.1.5 Printed Circuit Board (PCB) Analysis and Discrete Component Selection Figure 3.8 Example PCB temperature map. |
| 56 | 3.4.1.6 Heat Sinks (Air-Cooling) Figure 3.9 Example heat sinks. Figure 3.10 Example heat pipe heat sink. |
| 57 | 3.4.1.7 Liquid-Cooling Implementations Figure 3.11 Liquid-cooling implementations. |
| 58 | Table 3.1 Thermophysical Property Comparison: Water versus Air Table 3.2 FOM Ratio: Water-to-Air |
| 59 | Figure 3.12 External water cooling using an external CDU. |
| 60 | Figure 3.13 Schematic of (internal) closed-loop cooling. |
| 61 | Figure 3.14 Apple Power Mac G5 internal liquid-cooling system. Figure 3.15 Commercially available desktop internal liquid-cooling systems: (a) CoolIT Systems, (b) Asetek, Inc. |
| 62 | Figure 3.16 IBM zEnterprise EC12 processor (internal liquid) cooling system hardware (top) and cooling loop schematic (bottom). Figure 3.17 IBM water-cooled processor—Thermal conduction module. |
| 63 | Figure 3.18 IBM Power 775 Supercomputer—Water-cooled processors (Goth et al. 2012). |
| 64 | Figure 3.19 IBM Power 775 Supercomputer—Processor cold plate (Ellsworth et al. 2008). Figure 3.20 Fujitsu K Supercomputer—Modular LCU (Wei 2011). |
| 65 | Figure 3.21 Aluminum plate with embedded copper tube cold plate. |
| 66 | Figure 3.22 IBM Power 775 Supercomputer—Liquid-cooled memory. Figure 3.23 IBM System 360—Indirect (hybrid) air/water cooling. |
| 67 | Figure 3.24 Rear door heat exchanger. |
| 68 | Figure 3.25 Boundaries for thermal management. |
| 69 | Figure 3.26 Thermal control process. 3.5.1 Node-Level Management 3.5.1.1 ITE-Embedded Management Controller Subsystem |
| 70 | 3.5.1.2 Management Subsystem Communication Interfaces 3.5.1.3 Platform Hardware Connectivity |
| 71 | 3.5.2 Thermal Management Overview Figure 3.27 Example node thermal/power management subsystem. |
| 72 | Figure 3.28 Example 1U volume server thermal sensors. Figure 3.29 Example board sensor locations. |
| 73 | 3.5.2.1 Margin Versus Absolute Sensors 3.5.2.2 Sensor Aggregation 3.5.2.3 Sensor Averaging |
| 74 | 3.5.2.4 Thermal Sensors |
| 75 | Figure 3.30 Example temperature distribution on a CPU die. 3.5.2.5 Memory 3.5.2.6 Support Logic 3.5.2.7 Voltage Regulators (VR) |
| 76 | 3.5.2.8 Power Supplies 3.5.3 Add-In Modules and Miscellaneous Board Components 3.5.3.1 HDDs and SSDs |
| 77 | 3.5.3.2 Board Temperatures 3.5.3.3 Power 3.5.3.4 Compute Usage per (Time) Period 3.5.3.5 Fans |
| 78 | 3.5.3.6 Other Sensors |
| 79 | 3.5.4 Fan Speed Control and Design Figure 3.31 Fan zone mapping. |
| 80 | Figure 3.32 Example sensor mapping. |
| 81 | Figure 3.33 Fan speed versus system inlet temperature. |
| 82 | Figure 3.34 Example of rack level storage array airflow as function of IT inlet air temperature. |
| 83 | 3.5.4.1 Fan Zones Figure 3.35 Fan zone examples. |
| 84 | 3.5.4.2 Configuration, Workload, and Environment Table 3.3 Fan Zone Sensor Mapping |
| 85 | 3.5.4.3 System Tuning and Configurability 3.5.5 Future Integration with Data Center Management |
| 86 | 3.5.5.1 Integrated ITE-Data Center Control |
| 87 | 3.5.5.2 Evolution of Data Center Cooling Control |
| 89 | Figure 4.1 Server Dt projection at 25°C (77°F) server inlet temperature. |
| 90 | Figure 4.2 Server Dt projection at 35°C (95°F server inlet temperature. Figure 4.3 Server Dt projection at 45°C (113°F) server inlet temperature. |
| 91 | Figure 4.4 Dt expressed as cfm/kW at 25°C (77°F) server inlet temperature. |
| 92 | Figure 4.5 Future airflow trends based on ASHRAE 42U rack power trends at 20°C (68°F) Dt (90 cfm/kW [153 m3/kW·h]). |
| 94 | Figure 4.6 Open-delivery, raised-access floor option. Figure 4.7 Directly coupled: Hot-aisle containment. |
| 95 | Figure 4.8 Directly coupled: Cold-aisle containment. Figure 4.9 Directly coupled: Hot-exhaust chimney. Figure 4.10 Directly coupled: Inlet plenum. |
| 96 | Figure 4.11 Directly coupled: Rear-door heat exchanger. |
| 97 | Figure 4.12 Pressure implications of product choice and deployment strategies. |
| 98 | Figure 4.13 Pressure experienced by ITE in standard rack with perforated front and rear doors. 4.3.1 Data Center Pressure Effects on ITE |
| 99 | Figure 4.14 Server power consumption versus external pressure. |
| 100 | Figure 4.15 Component temperatures versus external pressure. |
| 101 | Figure 4.16 Server flow rate versus external pressure. |
| 102 | Figure 4.17 Rack airflow consumption: Stepwise thermal control versus closed loop. 4.3.2 Additional Server Data |
| 103 | Figure 4.18 Example of server control with and without hindering pressure. Figure 4.19 Six servers: Percentage flow rate change versus pressure. |
| 104 | 4.3.3 Top-of-Rack (TOR) Network Switches and Pressure Figure 4.20 TOR switch response to pressure. |
| 105 | 4.3.4 Hindering Pressures Typical of Passive Tight Containment |
| 106 | Figure 4.21 Active cooling device: Ducted switch cooler. Figure 4.22 Active cooling device: Rack cabinet fans. |
| 107 | 4.3.5 Implications of Helping Pressure 4.4.1 ITE Inlet Temperature Rate of Change |
| 108 | 4.4.2 Impact of Cooling Failure on ITE |
| 109 | Figure 4.23 Example transient response of a data center based on different cooling failures and assuming a 20°C (36°F) server Dt (Erden et al. 2014). 4.5.1 Electrostatic Discharge (ESD) |
| 111 | 4.5.2 Gaseous and Particulate Contamination |
| 112 | 4.5.2.1 ITE Filtration |
| 113 | 4.5.2.2 Heat Sink and Other Fouling |
| 114 | Figure 4.24 Example of expected increase in A-weighted sound power level with temperature. |
| 116 | Figure A.1 Example component with heat sink. |
| 118 | Figure A.2 Thermal characteristics of an example 1U heat sink. YCA stands for case to ambient thermal resistance. |
| 119 | Figure A.3 Parallel-plate heat sink. |
| 121 | Figure A.4 Heat sink optimization procedure. The SA subscript on the y- axis label stands for surface to ambient. |
| 122 | Figure A.5 DIMM with heat spreader. |
| 125 | Figure B.1 DIMM power versus bandwidth. Note: BW/DIMM = amount of transactions per time (BW) per DIMM. Figure B.2 DIMM cooling comparison. |
| 127 | Figure B.3 Cooling capability comparisons of different heat sinks. |
| 128 | Figure B.4 Improving memory cooling capability. |
| 131 | C.3.1 Description C.1.1.1 Power Trending C.1.1.2 Power Management C.3.2 Purpose C.1.2.1 Power Monitoring |
| 132 | C.3.3 Implementation and Impact C.2.1 Description C.2.1.1 Thermal Reporting |
| 133 | C.2.2 Purpose C.2.3 Implementation and Impact C.3.1 Description |
| 134 | C.3.2 Purpose C.3.3 Implementation and Impact C.4.1 Description |
| 135 | C.4.2 Purpose C.3.3 Implementation and Impact |
