Network Fabric Power Management Through Bandwidth Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network fabrics in computing systems are designed to handle worst-case scenarios, leading to excessive peak power consumption and increased circuit sizes, complexity, and inefficiency due to the mismatch between theoretical peak bandwidth and actual application bandwidth.
Innovation Solution
Implementing a bandwidth regulation circuit that predicts future power allocation needs and adjusts network power consumption based on estimated bandwidth requirements, using a centralized or hierarchal structure to manage power distribution among network switching circuits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the network fabric is designed to handle worst-case peak bandwidth conditions, then functional robustness and reliability are improved, but peak power consumption and circuit size increase excessively
Solution Approach 1:
The patent implements dynamic bandwidth allocation that adjusts power consumption levels based on actual traffic conditions. The system transitions from static worst-case design to dynamic adaptation, where power allocation matches real-time bandwidth requirements. This resolves the contradiction by maintaining reliability through adequate power allocation while avoiding excessive consumption during low-traffic periods.
Solution Approach 2:
The system changes power consumption parameters dynamically based on actual bandwidth utilization. By monitoring traffic patterns and adjusting power allocation accordingly, the system maintains functional robustness while optimizing power usage. This parameter adaptation allows the network fabric to handle worst-case conditions when necessary while consuming less power during normal operation.
2Reliability
If the network fabric is designed for peak bandwidth conditions, then worst-case performance is ensured, but actual application bandwidth efficiency decreases due to overdesign
Solution Approach 1:
The patent employs dynamic bandwidth allocation that adapts to actual traffic conditions rather than operating at fixed peak capacity. This dynamic adjustment ensures worst-case handling capability is available when needed while improving actual application bandwidth efficiency during normal operation by allocating resources based on real-time demands.
Solution Approach 2:
The network fabric implements self-service mechanisms where the system automatically monitors and adjusts its own power allocation based on actual bandwidth utilization. This self-regulation ensures worst-case capability is maintained while optimizing efficiency for actual application needs without external intervention.
3Speed
If fully buffered network switches are used to support unified memory space, then high bandwidth capability is achieved, but circuit complexity and power consumption increase
Solution Approach 1:
The patent implements dynamic resource allocation in network switches that adjusts buffering and bandwidth allocation based on actual traffic patterns. This dynamic approach maintains high bandwidth capability when needed while reducing circuit complexity and power consumption during lower-utilization periods by deactivating or scaling down unused resources.
Data Source
AI summary
An apparatus includes a system-on-chip that includes a plurality of agents configured to generate data transactions, a communication network configured to transfer transactions between two or more of the agents, a plurality of network switches, and a bandwidth regulation circuit. The network switching circuits may be coupled to the agents and to the network. One of the network switches may be configured to estimate a bandwidth need for transactions to be sent via the network switch in an upcoming window. The bandwidth regulation circuit may be configured to moderate power consumption of the network by determining a bandwidth budget using a network power budget for the upcoming time window, and determining a global bandwidth forecast using estimated bandwidth needs received from the network switches. The bandwidth regulation circuit may also be configured to allocate, using the global bandwidth forecast, the bandwidth budget among the network switches.