Multi-Layer Cache Pipelines for Higher Capacity at CPU Frequency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The cache capacity of CPUs is limited by frequency improvement and power consumption optimization, acting as a bottleneck for performance enhancement due to limitations in the pipeline stages of instruction execution and cache access.
Innovation Solution
A cache structure with at least two layer structures, each comprising a first and second cache unit, where cache access instructions enter different pipelines to access corresponding units in staggered sequences, expanding capacity without reducing frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If cache capacity is increased, then data hit rate improves, but processing frequency decreases and power consumption increases
Solution Approach 1:
The cache is divided into multiple independent layer structures (at least two layers), each with its own pipeline for accessing cache units. This segmentation allows parallel processing of cache access instructions across different layers, maintaining high processing frequency while increasing overall cache capacity through the combined storage of multiple layers.
Solution Approach 2:
The patent transitions from a single-layer cache structure to a multi-layer cache structure, adding a dimensional aspect to cache organization. Each layer operates independently with dedicated pipelines, effectively increasing cache capacity without proportionally increasing the complexity of the access path, thus maintaining processing frequency.
2Quantity of substance
If cache capacity is increased, then data hit rate improves, but power consumption increases
Solution Approach 1:
By segmenting the cache into multiple layers with independent pipelines, each layer can be activated independently based on access patterns. This reduces overall power consumption compared to a single large cache that must be continuously accessed, as only the necessary layers are actively processed at any given time.
Solution Approach 2:
The multi-layer structure enables partial action where only specific layers are accessed based on the needs of the processing task. Not all cache layers need to be fully utilized or accessed simultaneously, allowing the system to consume only the necessary power for the current operational requirements while maintaining the capacity benefits of a larger cache.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The present disclosure provides a cache, including at least two layer structures, each of the at least two layer structures includes: a first cache unit and a second cache unit, first cache access instructions of different layer structures enter different pipelines to access the first cache units in corresponding layer structures, and second cache access instructions of the different layer structures enter different pipelines to access the second cache units in corresponding layer structures.