Automated HBM Channel Identification for 3DIC Auto-Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge in 3DIC systems is the manual and error-prone process of identifying high-bandwidth memory (HBM) channels for auto-routing, which is complicated by high operating frequencies, long interconnect lengths, and the need for signal integrity and noise tolerance, making existing routers ineffective in automatically routing HBM channels.
Innovation Solution
An automated method and system that identifies HBM channels by determining candidate nets, calculating bounding boxes, and analyzing bump patterns to generate subchannels, reducing the need for manual input and improving routing quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual specification of HBM channel nets is performed, then routing accuracy can be maintained, but the process becomes tedious and error-prone
Solution Approach 1:
The system automatically identifies HBM channel nets by analyzing the netlist and detecting characteristic patterns such as bump arrays and channel configurations, eliminating the need for manual specification while maintaining routing accuracy through automated detection algorithms
Solution Approach 2:
The manual mechanical process of specifying nets is replaced with an automated computational system that uses algorithms to detect HBM channel characteristics in the netlist, transforming a labor-intensive manual task into an automated electronic identification process
2Adaptability or versatility
If existing conventional routers are used for HBM channel routing, then general routing functionality is available, but they cannot resolve the specific requirements of HBM channels
Solution Approach 1:
The routing system is customized with specific algorithms that detect and handle HBM channel characteristics such as bump patterns, channel orientations, and net configurations, providing specialized local processing for HBM channels while maintaining general routing capabilities for other nets
Solution Approach 2:
The routing process is segmented into distinct phases: automatic HBM channel identification, parameter extraction, and specialized routing, allowing the system to apply different processing strategies to HBM channels versus other routing tasks
3Productivity
If HBM channels are routed with high frequency and long interconnect length, then memory throughput is increased, but signal integrity and noise tolerance become problematic
Solution Approach 1:
The system performs preliminary analysis of HBM channel characteristics including bump patterns and channel configurations before routing, allowing optimization of routing parameters in advance to maintain signal integrity while achieving high throughput performance
Solution Approach 2:
The routing system adjusts various parameters such as interconnect width, layer selection, and via placement based on detected HBM channel characteristics to optimize both throughput and signal integrity for high-frequency operations
Data Source
AI summary
Methods and systems are described herein relate to automatic channel identification of high-bandwidth memory channels and subchannel generation. An HBM channel identification system may perform a sequence of operations to identify HBM channels within a netlist of an interposer: channel dimension prediction, channel bounding box prediction, channel orientation derivation, subchannel partition, and subchannel routing region creation. In one example, an HBM channel identification method includes identifying candidate nets within a netlist. A bounding box that includes one or more nets of the candidate nets is determined. Once the bounding box is determined, the orientation of the box is determined and used to determine a pattern of bumps within the bounding box. Finally, a subchannel is generated based on the pattern of bumps.