Dynamic Frequency Scaling and Dynamic Voltage Scaling
Last Updated : 01 Nov, 2019
'Frequency' in this context refers to the clock frequency or the frequency of operation of a CPU. So the term
Dynamic Frequency Scaling refers to the change of the clock frequency of the CPU during runtime. Now the above definition would instantly give rise to the question, Why do we need to do that? The answer to this question lies at the trade off between Performance and Power Consumption. We know that the
performance of a processor depends on
2 metrics, these are:
- CPU Response Time
- Throughput of the CPU
Performance metric is used for the determination of the performance of a Computer depends upon the application of the computer system. For example in case of Personal Computing Systems the
Response Time is very important and often is the sole basis for the determination of the performance of the computer. Whereas in case of
Servers the
throughput i.e. is the amount of work done in a given amount of time is much more important. But, for the purpose of this article when we say
'Performance' we will be concerning ourselves only with the
Response Time. On one hand we can increase the clock frequency of the CPU to reduce its response time and improve its performance but after a certain limit we need to also increase the voltage input to the CPU to maintain its stability at the high clock frequencies which in turn increases the Power Consumption and Heat Dissipation of the CPU, thereby shortening its lifespan. On the other hand we can reduce the clock frequency of the CPU below the standard values allowing us to undervolt the CPU and hence reduce the amount of Power Consumption by the CPU, but this has a negative impact on the CPU performance.
Performance {\propto} Clock Frequency Power Consumption {\propto} Clock Frequency
Now, Dynamic Frequency Scaling is a technique to balance the performance and Power Consumption. It refers to a continual variation of the clock frequency to optimize performance and Power Consumption of a CPU. Now the manner in which the CPU frequency is scaled is determined by the
frequency scaling algorithm used and the present CPU load. These frequency scaling algorithms are part of the
Kernel Code. Some of the most common Frequency Scaling Algorithms used in the Linux Kernel are:
- Performance: This Frequency Scaling Algorithm statically fixes the frequency of the CPU to its highest possible value. This increases the Power Consumption by the CPU
- Powersave: This Frequency Scaling Algorithm Statically fixes the frequency of the CPU to its lowest possible value. This takes it toll on the performance of the system.
- Conservative: This Frequency Scaling Algorithm adjusts the frequency of the CPU to a certain minimum possible value so as to keep the CPU load below a certain percentage. This algorithm tries to optimize the power consumption while keeping the power consumption.
Similar Reads
Horizontal and Vertical Scaling | System Design In system design, scaling is crucial for managing increased loads. This article explores horizontal and vertical scaling, detailing their differences. Understanding these approaches helps organizations make informed decisions for optimizing performance and ensuring scalability as their needs evolveH
8 min read
Overview of Scaling: Vertical And Horizontal Scaling Given architecture is an example of a client-server based system. In this, there is a client who sends requests to the server and then the client receives a response from the server accordingly but when the number of users/clients increases, the load on the server increases enormously which makes it
4 min read
Horizontal and Vertical Scaling In Databases Scaling alters the size of a system. In the scaling process, we either compress or expand the system to meet the expected needs. The scaling operation can be achieved by adding resources to meet the smaller expectation in the current system, by adding a new system to the existing one, or both. Type
3 min read
Difference between scaling horizontally and vertically for databases Scaling :It can be outlined as changing the size of something, for instance scaling the business. It is even same within the context of databases.There are two types of scaling :1. Horizontal scaling :Horizontal scaling means we scale by adding additional machines to our existing bunch of resources.
1 min read
Manual Scaling vs. Auto-Scaling in System Design In system design, scaling is key to handling variable loads efficiently. Manual scaling allows engineers to adjust resources based on expected demand, offering control but requiring oversight. Auto-scaling automates this process, dynamically adapting to real-time demand, improving resource efficienc
3 min read