Co-Packaged ASIC and HBM Thermal Control With Activity Shaping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern routing and switching systems face challenges in cooling co-packaged ASIC and HBM dies due to thermal cross-contamination and differing thermal resistance, leading to inefficient fan management and increased power consumption.
Innovation Solution
Implementing a temperature control method that adjusts controller traffic and fan speed based on target temperatures and activity levels, using a window algorithm to allow temporary HBM temperature exceedance, a shaper algorithm to limit activity, and continuously analyzing fan speed to ensure adequate cooling while reducing power usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Temperature
If fan speed is increased to cool both ASIC and HBM dies, then cooling effectiveness is improved, but power consumption increases
Solution Approach 1:
The patent divides the cooling control into separate segments for ASIC die and HBM die, with independent temperature monitoring and control algorithms. Each die type has its own temperature thresholds and cooling requirements that are managed separately rather than using a single unified cooling approach.
Solution Approach 2:
The fan speed is dynamically adjusted based on real-time temperature readings from both ASIC and HBM dies. The control algorithm continuously monitors temperatures and modifies fan speed accordingly, transitioning between different operating states (e.g., low-speed, medium-speed, high-speed) to match actual cooling需求的.
2Reliability
If fan speed is increased to ensure adequate cooling, then temperature control is improved, but power consumption increases
Solution Approach 1:
The system implements continuous temperature feedback from both ASIC and HBM dies to the control algorithm. Temperature sensors monitor the dies and feed this information back to adjust fan speed in real-time, ensuring reliable temperature control while avoiding unnecessary high-power operation when cooling demand is low.
Solution Approach 2:
The control algorithm changes operational parameters (fan speed levels) based on temperature conditions. Different temperature ranges trigger different fan speed parameters, allowing the system to adapt its power consumption to actual thermal conditions rather than operating at constant high power.
3Productivity
If activity is increased to improve system performance, then productivity is improved, but heat generation increases
Solution Approach 1:
The system dynamically balances activity levels with thermal conditions. When temperatures are within acceptable ranges, the system can operate at higher activity levels for improved performance. When temperatures approach thresholds, the system dynamically reduces activity or increases cooling to maintain safe operating conditions.
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
Effectively manages cooling for both ASIC and HBM dies, reducing power consumption and maintaining device integrity by optimizing fan speed and activity allocation, while ensuring temperatures remain within safe limits.
Implementation Method 1
a fan to cool the co-packaged die
Implementation Method 2
Heatsinks, fans, and other techniques can be used to reduce or eliminate this heat
Data Source
AI summary
Techniques for temperature control for multiple dies in an element. A temperature of a first die is measured, in an element including the first die and a second die. The second die includes at least a portion of a controller. The temperature of the first die is changed by adjusting activity, from the second die to the first die, based on a target temperature for the first die and the measured temperature for the first die.