Unified Translation Queue for Multi-Mode Address Translation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current computing systems face inefficiencies and increased complexity due to the need to support multiple address translation modes, leading to higher chip area and power consumption, as each translation mode requires dedicated hardware and unique interfaces, resulting in suboptimal address translation latency and performance.

Innovation Solution

A unified memory address translation system is introduced, featuring a translation queue module, a translation cache interface, and a unified translation walk address generation module that share components across different translation modes, enabling efficient address translation by processing multiple requests in various modes, such as hashed page table, segment table, and radix requests, while reducing redundant hardware and logic.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If dedicated hardware is provided for each address translation mode, then address translation functionality is supported, but chip area and power consumption increase

Engineering Contradiction:
Improveaddress translation mode supportVSAvoidchip area
Core Design Contradiction:
Adaptability or versatilityVSArea of stationary object

Solution Approach 1:

The patent implements a unified translation queue module that serves multiple address translation modes (HPT, segment table, radix, and nested radix) through a single hardware structure. This multi-functional design eliminates the need for separate dedicated hardware for each translation mode, thereby reducing chip area while maintaining support for all required translation functionalities.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent combines multiple translation mode handlers into a single unified translation queue module. By merging the functionality of HPT translation, segment table translation, radix translation, and nested radix translation into one shared hardware structure, the design reduces overall chip area consumption compared to having separate dedicated hardware for each mode.

Inventive Principle:
Principle #5Merging (Combining)

2Adaptability or versatility

If dedicated hardware is provided for each address translation mode, then address translation functionality is supported, but power consumption increases

Engineering Contradiction:
Improveaddress translation mode supportVSAvoidpower consumption
Core Design Contradiction:
Adaptability or versatilityVSUse of energy by stationary object

Solution Approach 1:

The unified translation queue module is designed to handle multiple address translation modes through a single hardware structure that can be dynamically configured. This multi-functional approach ensures that only the necessary translation logic is active at any given time, reducing overall power consumption compared to having all dedicated hardware structures continuously powered.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

By merging multiple translation mode handlers into one unified module, the patent reduces the total number of active hardware components needed at any moment. This consolidation allows for more efficient power management, as the unified structure can share resources and reduce standby power consumption across different translation modes.

Inventive Principle:
Principle #5Merging (Combining)

3Adaptability or versatility

If multiple dedicated translation hardware structures are used, then different translation modes are supported, but address translation latency increases

Engineering Contradiction:
Improvetranslation mode diversityVSAvoidaddress translation latency
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The unified translation queue module is segmented into multiple independent translation walk queues, each capable of handling different translation modes simultaneously. This segmentation allows parallel processing of translation requests for different modes, reducing overall latency compared to a single sequential processor that would have to handle all modes one at a time.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements a dynamic arbitration mechanism that can allocate translation resources adaptively based on the current mix of translation requests. This dynamic resource allocation optimizes the processing of different translation modes in real-time, minimizing idle time and reducing overall address translation latency compared to static dedicated hardware configurations.

Inventive Principle:
Principle #15Dynamics

4Adaptability or versatility

If separate hardware and interfaces are provided for each translation mode, then translation functionality is complete, but device complexity increases

Engineering Contradiction:
Improvetranslation mode coverageVSAvoidhardware complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The unified translation queue module serves as a universal interface that handles multiple address translation modes through a single standardized entry point. This universal design simplifies the overall hardware architecture by eliminating the need for separate interfaces and control logic for each translation mode, thereby reducing device complexity while maintaining full translation mode coverage.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS11734188B2Unified translation miss queue for multiple address translation modes
Publication Date: 2023.08.22 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US11734188B2 patent drawing
  • US11734188B2 patent drawing
  • US11734188B2 patent drawing

AI summary

A unified memory address translation system includes a translation queue module configured to receive different modes of translation requests for a real address (RA) of a physical memory. A translation cache (XLTC) interface is configured to receive successful translation results for previous requests for an RA and provide the previous successful translation result to the translation queue module. A plurality of page table entry group (PTEG) search modules are coupled to the translation queue module. A unified translation walk address generation (UTWAG) module is configured to provide a translation support for each mode of the different modes of translation request. A memory interface is coupled between the UTWAG and the physical memory.