Multi-Slot Cache Layout for Direct Host Data Transfer

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Data in cache memory is often in an inefficient format, leading to slower data transfers and increased computing resource usage due to the need for additional processing such as Scatter-Gather (SGL) operations and inefficient decompression/compression for local and remote replication.

Innovation Solution

Utilizing multiple cache slots in storage arrays to store data in different formats, including compressed, native, and host transfer formats, allowing direct and efficient data transfer without the need for additional formatting.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If data is stored in a single cache slot in the format from storage array memory, then data transfer requires additional processing (SGL operations, decompression/compression), but using multiple cache slots with different data formats increases cache memory structure complexity

Engineering Contradiction:
Improvedata transfer efficiencyVSAvoidcache memory structure
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The cache memory is segmented into multiple cache slots, each dedicated to storing data in a specific format (e.g., compressed format, native format, host transfer format). This segmentation allows data to be stored in the most appropriate format for its intended operation, eliminating the need for format conversion and SGL processing during data transfers.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different cache slots are assigned different data formats based on their specific function or operation type. For example, one cache slot may store data in compressed format for efficient storage, while another stores data in native format for rapid access. This local quality approach optimizes performance for specific operations without requiring the entire cache to handle all format conversions.

Inventive Principle:
Principle #3Local quality

2Adaptability or versatility

If data is converted between different formats during cache operations, then data can be accessed in the required format, but additional CPU resources and processing time are consumed

Engineering Contradiction:
Improvedata format flexibilityVSAvoidCPU resource usage
Core Design Contradiction:
Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

Data is pre-formatted and stored in multiple cache slots in the formats needed for different operations before any data transfer or access occurs. This preliminary action eliminates the need for real-time format conversion during data operations, thereby reducing CPU resource usage and processing time.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates copies of data in different formats and stores them in separate cache slots. Instead of converting data formats during operations, pre-formatted copies are already available for immediate use, significantly reducing the computational overhead associated with format conversion.

Inventive Principle:
Principle #26Copying

3Ease of operation

If Scatter-Gather processing is used for data transfers, then data can be assembled from multiple sources, but host read response times increase and throughput decreases

Engineering Contradiction:
Improvedata assembly capabilityVSAvoidhost read response time
Core Design Contradiction:
Ease of operationVSSpeed

Solution Approach 1:

Data is pre-assembled and organized in the required formats in appropriate cache slots before host access requests are processed. This preliminary assembly eliminates the need for Scatter-Gather processing during host read operations, thereby reducing response times and increasing throughput.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The cache memory system acts as an intermediary between the storage array and the host, pre-processing and organizing data in the required formats before transfer. This intermediary function eliminates the need for Scatter-Gather operations at the host level, improving data transfer speed and host read response times.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS12536102B2Purposed data transfer using multiple cache slots
Publication Date: 2026.01.27 DELL PROD LP
  • US12536102B2 patent drawing
  • US12536102B2 patent drawing
  • US12536102B2 patent drawing

AI summary

An example methodology includes provisioning memory of a storage array with cache memory including a plurality of cache slots and reading data from the memory of the storage array into a first cache slot of the plurality of cache slots, wherein the data in the first cache slot is in a first data format, the first data format being the format of data in the memory of the storage array. The method also includes converting the data in the first cache to a second data format distinct from the first data format and placing the data in the second data format into a second cache slot of the plurality of cache slots.