Client Address Translation Cache for Faster MMU Bypass
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high data rate and translation request rate in processing systems result in many MMU cache misses, leading to slow translation production due to the inefficiency of translation lookaside buffers (TLBs) and the need for manual page table walks in dynamic random access memory (DRAM).
Innovation Solution
Implementing a translation extension unit with a client address translation cache that stores a cached group of address translations, allowing the system to return physical addresses without transmitting requests to the memory management unit when the virtual address matches a cached entry, and updating the cache with invalidation messages to maintain validity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the system uses MMU with TLB for address translation, then translation requests can be processed, but translation speed decreases due to cache misses and manual page table walks in DRAM
Solution Approach 1:
The invention segments the address translation function by introducing a dedicated translation extension unit (TEU) that operates independently from the main MMU. Each TEU is associated with a specific client and maintains its own translation cache, separating the translation workload into parallel units that can process requests simultaneously without blocking the main MMU.
Solution Approach 2:
The invention creates copies of the translation cache functionality at the client side through the translation extension units. Instead of relying solely on the central MMU's TLB, each TEU maintains a local cache of address translations, effectively copying the caching function to reduce dependency on the central MMU and enable faster local translation.
2Productivity
If the system increases data rate and translation request rate, then processing efficiency improves, but MMU cache misses increase leading to slower translation production
Solution Approach 1:
The invention divides the translation processing into multiple independent segments by creating separate translation extension units for different clients. Each TEU handles translation requests for its associated client independently, allowing parallel processing that reduces the burden on the central MMU and maintains higher cache hit rates even under increased request volumes.
Solution Approach 2:
The invention adds a new dimension to the translation architecture by introducing client-side translation extension units that operate in parallel with the central MMU. This dimensional expansion allows translation requests to be processed at multiple levels simultaneously, improving overall throughput while maintaining reliability through distributed caching.
Data Source
AI summary
An address translation system, method and device including a translation extension unit associated with a client and having a client address translation cache storing a client cache of the address translations. When virtual address of a translation request from the client is the same as the virtual address of an address translation of the client cache, the translation extension unit returns the physical address that is mapped to the virtual address according to the address translation to the client without transmitting the translation request to a memory management unit that maintains a page table of all the translations and a translation lookaside buffer with a subset of the translations.