A method for intelligent control of fans

Abstract

The invention provides a method for intelligently regulating and controlling a fan. The method specifically includes;firstly, acquiring height size of a single fan structure, and regionalizing server nodes according to the height size of the fan structure; acquiring temperature condition of the server nodes in groups; outputting PWM (pulse-width modulation) signals for regulating and controlling the rotating speed of the fan by the nodes according to above temperature condition; comparing and judging PWM duty ratios; and switching to the high PWM duty ratio for controlling the fan. Compared with the prior art, the method has the advantages that high efficiency and low noise of regulation and controlling of multiple server nodes are guaranteed, intelligence of fan regulation and controlling is achieved, and the method is higher in reliability and stability.

Description

technical field [0001] The invention relates to the technical field of computer communication, in particular to a high-efficiency and low-noise fan intelligent control method that ensures multi-server node control. Background technique [0002] In today's server system, it is more and more common to adopt the topology of multi-node cluster distribution, that is, multiple server nodes are gathered together to form a server cluster. The computing tasks of the server cluster are getting higher and higher, and its ability to respond to external cooling is getting better Fast, while completing high-performance calculations with large amounts of data, higher requirements are put forward for the peripheral fan control design. Fan control needs to take into account real-time response capability, power consumption, and noise control; the real-time response capability of fan control is required. When the temperature of the server node rises rapidly, the fan can be controlled to speed ...

AI Technical Summary

Problems solved by technology

[0003] The current heat dissipation design for server multi-node systems can only take into account the overall or constant mode, that is, all nodes are taken into consideration. When the overall power is low, the fan speed is reduced, or the fan is fixed at a high speed of 80% and remains constant; the same fan is Multiple server nodes provide cooling air volume, and the unbalanced server computing load leads to low efficiency of unified cooling; the overall cooling strategy cannot guarantee effective and timely cooling when a single node has a large load capacity, and the constant mode cooling strategy cannot dynamically change the wind speed in real time , it is easy to cause waste of energy, and the noise at high speeds cannot be effectively controlled

Images