Switch Point Network for AXI Write Ordering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In producer-consumer systems using the AXI protocol, existing methods for ordered write transactions result in high latencies, reduced bandwidth, and increased buffering requirements at the requester node due to the need for adhering to AXI ordering rules, especially when dealing with interleaved write transactions across multiple target devices.
Innovation Solution
Implementing a network of interconnected switch points that control the flow of write requests and responses, allowing the switch point interconnection network to manage AXI ordering rules, thereby reducing the processing overhead and buffering requirements at the requester node by handling ordering rules within the interconnection at a destination target bifurcation point.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AXI ordering rules are enforced at the requester node, then write transaction ordering is maintained, but latency increases and bandwidth decreases
Solution Approach 1:
The patent extracts the ordering enforcement function from the requester node and relocates it to the interconnection network itself. The network now monitors and enforces AXI ordering rules independently, allowing the requester to issue writes without stalling, thus reducing latency while maintaining ordering guarantees through network-level control.
Solution Approach 2:
The interconnection network acts as an intermediary between the requester and target devices, taking on the responsibility of enforcing ordering rules. This mediator approach allows the requester to operate independently while the network ensures proper ordering is maintained across multiple targets, resolving the contradiction between ordering reliability and latency.
2Reliability
If AXI ordering rules are enforced at the requester node, then write transaction ordering is maintained, but bandwidth is reduced
Solution Approach 1:
By extracting the ordering enforcement responsibility from the requester node to the interconnection network, the requester can continuously issue write transactions without stalling. This increases bandwidth utilization while the network maintains ordering guarantees through its centralized control mechanism.
Solution Approach 2:
The interconnection network as an intermediary enables higher bandwidth by allowing parallel write issuance from the requester while managing ordering constraints centrally. This eliminates the need for the requester to serially enforce ordering, thus increasing throughput while maintaining reliability.
3Reliability
If AXI ordering rules are enforced at the requester node, then write transaction ordering is maintained, but buffering requirements increase
Solution Approach 1:
The patent removes the need for large buffering structures at the requester node by transferring ordering enforcement to the network. The requester can use smaller buffers since it no longer needs to hold writes pending completion at previous targets, as the network will manage the ordering independently.
Solution Approach 2:
The interconnection network as a mediator assumes the buffering and ordering management responsibility, allowing the requester to reduce its own buffering requirements. The network can implement ordering logic with its own buffering mechanisms, decoupling the requester's buffer size from the ordering enforcement needs.
4Productivity
If buffered writes are issued to multiple target devices, then throughput is improved, but maintaining ordering becomes more complex
Solution Approach 1:
The patent extracts the complex ordering management logic from the requester node and relocates it to the interconnection network. This allows the requester to issue buffered writes to multiple targets simultaneously, improving throughput, while the network handles the complexity of ensuring proper ordering is maintained across all targets.
Solution Approach 2:
The interconnection network serves as an intermediary that manages the complexity of ordering across multiple targets. It receives buffered writes from the requester, tracks completion status across different targets, and ensures proper ordering is maintained, thereby reducing the requester's complexity while maintaining throughput.
Data Source
AI summary
Ordered write transactions from requester devices to multiple target devices are controlled using switch point networks. The requester device and the multiple target devices for the write transactions are coupled to a network of interconnected switch points. Write requests are generated for a plurality of parcels associated with a block of data to be written. The write requests have a particular order associated with an order in which the parcels are to be written, and these write requests are provided to the switch point interconnection network in the particular order. At least one of the switch points is then used to control the flow of write requests to the multiple target devices such that the particular order is maintained. In one embodiment, the target devices are memory devices, and the particular order is based upon the AXI (Advanced eXtensible Interface) protocol.