Clock signal

A clock signal or the system clock (English clock signal, clock or system clock period) is a binary signal which is used to coordinate the actions of multiple circuits, especially the synchronization of flip-flops within complex digital systems in digital technology. Depending on the application, the clock signal may be repeated with a fixed frequency or may also be aperiodic.

General

If a periodic timing signal is present, it is used by an oscillator such as a quartz oscillator generated. A conventional oscillation circuit for generating clock signals, the circuit Pierce.

Clock generation

The clock signal oscillates between the two logic levels as shown in the adjacent sketch with H for high and L for low. The clock signal is characterized by the period of the oscillation, or with its inverse, the frequency and the duty cycle.

To clock generating a square clock, quartz oscillators are used today in electronic circuits in addition to the quartz crystals together with the drive circuit. The advantages of these devices are the low tolerance of the frequency generated and the high stability of frequency over the permissible temperature range, the resistance to aging of the device and the permissible range for the operating voltage of the devices.

Modern processors and their auxiliary blocks on the motherboard of a computer need several different clock signals, as for example the CPU with a much higher clock runs as external interfaces. Frequencies are dynamically switched depending on the operating situation, especially for the purpose of energy conservation Even within the CPU. For the provision of such a variety of clock signals typically a master oscillator is responsible, which derives all required frequencies from an even higher crystal frequency has its own frequency divider.

CPUs and GPUs are available in versions for specific frequency ranges for which they were tested in manufacturing. Since these areas are designed to be more conservative security audience, many users - especially from the computer game area - to push the limits. Many processors offer for this overclocking to specific control registers, on the one a divider factor - can set that indicates how the operating frequency to be divided from the clock speed - the multiplier. However, such settings are risky, since, in extreme cases can happen to data loss or destruction of the processor. Therefore, there are companies that offer as part of its service offering such overclocked computer or graphics cards, which they have previously tested for adequate reliability.

Clock signal for integrated circuits

The circuits that use the clock signal for synchronization may vary according to type either during the rising or the falling edge of the signal is active ( DDR both edges are used ), this is referred to edge trigger or edge trigger. Data sheets and diagrams, the clock signal is commonly referred to as CLK. Using clock gating can be switched off in integrated circuits, the clock selectively for certain, currently not needed circuit parts in order to reduce the average power consumption.

The most complex integrated circuits require a clock signal to synchronize different parts of the chip and to compensate for gate delay. As these devices become more complex, the delivery of accurate and everywhere the same clocks to all circuits is becoming increasingly difficult. The prime example of such complex chips are microprocessors, the central components of modern computers. For the transistors used, the frequency is often specified to a gain of a small signal, is possible.

Speeds for computer processors

With the processor clock frequency is used to specify the speed at which can be processed in computers the data. It is expressed in Hertz ( Hz). Since the frequency of modern processors is several billion hertz, the numbers often with the help of intentions as Giga (G ) for billions or mega (M) for millions abbreviated ( in embedded systems but also partly kHz are common). For example, means a frequency of 1 GHz, a clock period of a nanosecond.

The data throughput of a processor arising from its clock and data rate of its connection to the main memory. The processing power (measured, for example, MIPS or flops) is dependent not only on frequency but also on the overall architecture of the processor. Even with processors that use the same instruction set, can at the same clock rate shows serious processing power differences, the cause of, for example, in the IPC rate (IPC: Instructions per cycle, instructions per clock ), firm-specific characteristics ( for example SIMD extensions ) or may be due to the aforementioned memory bandwidth. The IPC rate indicates how many instructions a processor can execute simultaneously through parallelization. The processor with higher IPC rate therefore creates more arithmetic operations per clock and therefore expects faster.