Rack Cooling for Computing Devices via Offset Shelves
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data centers housing large numbers of computing devices face cooling challenges due to heat density, power usage, and inefficiencies in traditional cooling methods, leading to hot spots and reduced reliability of mining rigs and other CPU/GPU-based systems.
Innovation Solution
A system and method that configures computing devices in racks to exhaust heat in a coordinated pattern, with fan speeds managed based on location to improve airflow, using gradient patterns and sloped positions of devices and racks to enhance cooling efficiency, and includes a control module to dispatch instructions for optimal fan speed settings and operating parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Temperature
If traditional cooling methods (additional large external fans, external cooling units) are used to improve cooling of computing devices in data centers, then cooling effectiveness is improved, but energy consumption increases significantly
Solution Approach 1:
The system implements periodic action by cycling fan speeds through different patterns (e.g., high-speed phase, medium-speed phase, low-speed phase) over time. Each computing device alternates between different operational states, creating a temporal distribution of cooling demand that reduces peak energy consumption while maintaining adequate cooling effectiveness.
Solution Approach 2:
The system changes operational parameters by dynamically adjusting fan speeds based on temporal patterns and spatial positions. Instead of maintaining constant high fan speeds, the system varies fan rotation speeds through different phases and patterns, optimizing the balance between cooling effectiveness and energy consumption.
2Temperature
If additional large external fans are added to increase airflow across computing devices, then cooling effectiveness is improved, but device complexity and system cost increase
Solution Approach 1:
The system implements self-service by utilizing the existing built-in cooling fans of computing devices rather than adding external cooling infrastructure. The management system coordinates and controls the fan speeds of these self-contained cooling components, enabling the devices to cool themselves through coordinated operation without requiring additional external cooling equipment.
Solution Approach 2:
The system applies universality by making the built-in cooling fans serve multiple functions: they provide both the primary cooling mechanism and act as controllable components for implementing complex cooling patterns. The same fan hardware performs both local device cooling and contributes to overall data center thermal management when coordinated across multiple devices.
3Temperature
If computing devices are positioned far from exhaust vents to optimize airflow patterns, then cooling efficiency is improved, but some devices suffer from reduced airflow and hot spots
Solution Approach 1:
The system applies local quality by assigning different fan speed patterns and cooling strategies to different spatial locations within the data center. Devices closer to exhaust vents receive different operational instructions than devices farther away, with local adjustments to fan speeds and cooling patterns that account for position-specific airflow characteristics, ensuring adequate cooling across all locations.
Solution Approach 2:
The system implements dynamics by continuously adjusting fan speeds and cooling patterns based on real-time conditions and device positions. Rather than static positioning and fixed fan speeds, the system dynamically modifies operational parameters to respond to changing thermal conditions, airflow patterns, and device locations, preventing hot spots and maintaining cooling uniformity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach improves cooling efficiency, reduces energy consumption, and prevents hot spots by optimizing airflow and heat management within data centers, thereby enhancing the reliability and longevity of computing devices.
Implementation Method 1
a cooling fan configured to move air from a cold aisle into a computing device positioned on the shelf and from the computing device to a hot aisle
Implementation Method 2
The computing device is configured to exhaust heat in a first direction that is parallel to other computing devices on the rack
Data Source
AI summary
Systems and methods for cooling large numbers of computing devices in a data center are disclosed. The devices are positioned on shelves in racks and are connected via network switches. The computing devices are oriented so that their cooling fans all exhaust waste heat to one side of the rack into a hot aisle, and each position is offset from positions on neighboring shelves above and below the current shelf. The racks can be offset, or the positions within the racks can be offset. Each computing device on the rack can also be angled horizontally and or vertically to improve airflow. Multiple racks can be configured together into sloped regular polygons, with the interior forming the hot aisle, and deflectors can be installed to further improve airflow.