Translation Lookaside Buffer Resource Reservation for QoS Prioritization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing virtual memory address translation systems face performance degradation due to resource competition between different quality-of-service levels in translation lookaside buffers, where lower quality-of-service requests can occupy resources, delaying higher quality-of-service requests.

Innovation Solution

Reserving translation lookaside buffer slots and finite state machines for specific quality-of-service levels ensures that higher quality-of-service requests are prioritized and processed without being delayed by lower quality-of-service requests, preventing resource competition and maintaining performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If translation lookaside buffer resources are shared across all quality-of-service levels, then resource utilization is improved, but performance of high quality-of-service requests deteriorates due to competition with low quality-of-service requests

Engineering Contradiction:

Improveresource utilizationVSAvoidperformance of high quality-of-service requests

Core Design Contradiction:

Quantity of substanceVSProductivity

Solution Approach 1:

The translation lookaside buffer resources are segmented into separate pools for different quality-of-service levels. High quality-of-service requests are allocated dedicated resources that are isolated from low quality-of-service requests, preventing resource competition while maintaining high resource utilization within each segment.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different quality-of-service levels are assigned different resource allocation policies and priorities. High quality-of-service requests receive preferential treatment with guaranteed resource availability, while low quality-of-service requests use remaining resources, creating localized optimization for each service level.

Inventive Principle:

Principle #3Local quality

2Productivity

If translation lookaside buffer resources are reserved for specific quality-of-service levels, then performance of high quality-of-service requests is improved, but overall resource utilization deteriorates due to resource fragmentation

Engineering Contradiction:

Improveperformance of high quality-of-service requestsVSAvoidoverall resource utilization

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The resource allocation system dynamically adjusts the distribution of translation lookaside buffer resources based on current workload conditions. When high quality-of-service requests are present, dedicated resources are allocated; when idle, resources can be reallocated or shared, maintaining high performance while improving overall utilization.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

Translation lookaside buffer resources are designed to serve multiple quality-of-service levels through a unified management structure. The same physical resources can be allocated to different service levels depending on demand, providing multi-functionality that prevents fragmentation while ensuring high-quality service when needed.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Quantity of substance

If low quality-of-service requests occupy translation lookaside buffer resources, then resource utilization is improved, but latency of high quality-of-service requests increases

Engineering Contradiction:

Improveresource utilizationVSAvoidlatency of high quality-of-service requests

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

Low quality-of-service requests are extracted from the shared resource pool and placed in a separate processing queue or allocated to dedicated lower-priority resources. This prevents them from occupying translation lookaside buffer slots that could be used by high quality-of-service requests, eliminating latency interference while maintaining resource utilization through separate pools.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

High quality-of-service requests are given preliminary access to dedicated translation lookaside buffer resources before low quality-of-service requests can occupy general resources. This advance allocation ensures that when high-priority requests arrive, resources are already reserved for them, preventing latency increases due to later competition.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS10114761B2Sharing translation lookaside buffer resources for different traffic classes

Publication Date: 2018.10.30 ADVANCED MICRO DEVICES INC

AI summary

Techniques are provided for managing address translation request traffic where memory access requests can be made with differing quality-of-service levels, which specify latency and/or bandwidth requirements. The techniques involve translation lookaside buffers. Within the translation lookaside buffers, certain resources are reserved for specific quality-of-service levels. More specifically, translation lookaside buffer slots, which store the actual translations, as well as finite state machines in a work queue, are reserved for specific quality-of-service levels. The translation lookaside buffer receives multiple requests for address translation. The translation lookaside buffer selects requests having the highest quality-of-service level for which an available finite state machine is available. The fact that finite state machines are reserved to particular quality-of-service levels means that if all such finite state machines for a particular quality-of-service level are used by pending translation requests, then the translation lookaside buffer does not accept more translation requests for that quality-of-service level.