Dynamic Cache Line Compression at Networking Stack

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In computing systems, the speed mismatch between processor cores and cache links leads to performance degradation due to limited link speed, resulting in increased latency and reduced bandwidth in cache line transfers.

Innovation Solution

Dynamic compression of cache lines at the lowest level of the networking stack, specifically within the framer and data link layer, based on various parameters, to optimize bandwidth and latency without increasing complexity or compromising security.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If data is transmitted at link speed, then bandwidth is limited, but latency increases and performance degrades

Engineering Contradiction:

Improvecache line transfer speedVSAvoidtransfer latency

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent applies compression algorithms to change the parameter of data representation, transforming cache line data into a compressed format that occupies less bandwidth. This allows more data to be transmitted through the fixed-speed link in the same time period, effectively increasing transfer speed and reducing latency without changing the physical link characteristics.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If compression is applied to cache lines, then bandwidth efficiency improves, but system complexity increases

Engineering Contradiction:

Improvebandwidth utilizationVSAvoidcompression system complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The compression system is designed to be self-regulating, automatically adjusting compression levels based on data patterns and link conditions. The system monitors its own performance and dynamically optimizes compression parameters without external intervention, reducing the need for complex control mechanisms while maintaining high bandwidth utilization.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent implements dynamic compression that adapts to varying data characteristics and link conditions. Rather than using a fixed compression approach, the system dynamically selects compression strategies based on real-time conditions, allowing it to maintain high efficiency while avoiding unnecessary complexity for different data types and transmission scenarios.

Inventive Principle:

Principle #15Dynamics

3Productivity

If compression algorithms are applied, then transfer efficiency improves, but security against side-channel attacks may be compromised

Engineering Contradiction:

Improvetransfer efficiencyVSAvoidsecurity

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent applies different compression strategies to different portions of cache line data based on their security sensitivity. Critical security-related data portions are handled with compression algorithms that maintain security properties, while non-critical portions use more aggressive compression. This localized approach allows high transfer efficiency overall while maintaining security where required.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS11792303B1Fast clear memory of system memory

Publication Date: 2023.10.17 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Various embodiments are provided herein for compressing data in latency-critical processor links of a computing system in a computing environment. One or more cache lines may be dynamically compressed at a lowest level of a networking stack based on one or more of a plurality of parameters prior to transferring a single-cache line, where the networking stack includes a framer and a data link layer.