Many embedded processors claim to have the lowest power consumption. But in fact no single component can maintain the lowest power consumption in all applications, because the definition of low power consumption is related to the application environment, and the chip design suitable for one application is likely to bring problems to another application. Most portable applications define low power consumption based on battery life. These applications have a wide range of functions and operating modes are ever-changing. To meet the application's power requirements, telecommunications system components must handle the required number of channels within the power budget while dissipating power through the package and board to ensure that components remain within the specified temperature range; in addition, these foundations Facility applications also place great emphasis on power consumption under maximum load conditions. Therefore, in order to meet power requirements, DSP vendors will select the most appropriate component process, circuit design, voltage and frequency operating point, and overall architecture for the target application.
Power saving technology
DSP vendors have many techniques that can be used to reduce power consumption and achieve performance goals, including:
â— Choose an appropriate process;
â—Crystal design technology;
â— Select the correct operating frequency and voltage;
â— Choose the right architecture, including integration, memory architecture and arithmetic processing unit;
â— Use a highly efficient package to ensure that components remain within the specified operating temperature range.
Power source
Regardless of the application, component power consumption includes the following sources:
Leakage power
The leakage current of the component is fixed and is independent of processor operation or operating frequency, but varies with process, operating voltage, and temperature. The leakage power of the low geometry process will increase exponentially with voltage and temperature.
Clocking power
The clock power of the component is proportional to the clock frequency. Most of the wafer area of ​​high-integration components is used for synchronous components such as memory or scratchpads. If the clock architecture is poorly designed, the power consumption will remain the same regardless of the actual workload of the components.
Operating power consumption (acTIve power)
Relating to the actual system function that the component was performing at the time.
In addition to the above sources, component power consumption is also affected by two major factors:
Component current
The higher the component current, the faster the battery power is consumed, sometimes exceeding the power budget range, causing the supply voltage to drop, causing the component to leave the normal operating area and causing errors.
Component/system temperature rise
If the component cannot be effectively dissipated, its temperature may exceed the rated range and cause an operation error.
The following optimization techniques address the various power consumption issues described above in different ways.
Choose the appropriate process
To optimize the performance and power consumption of different applications, Texas Instruments offers a variety of process types. For example, TI's 130nm low-leakage process has almost no leakage current at 1.5V operation, and most of the DSP is idle. For portable applications in the state, this low leakage process can help them save power. Another high-performance process has a large leakage current, but it can operate at 1.2V. The components of this process can achieve twice the MHz performance of the low leakage process. In infrastructure applications that focus on fully-acTIve power, this high-performance process is far more competitive than low-leakage processes for two reasons: First, the operating frequency of the low-leakage processing unit is only Half of the high-performance process, which means that the number must be doubled to provide the same performance, but this will lead to higher component costs. Secondly, since the power consumption is proportional to the square of the voltage, the operating power consumption of the high-efficiency process is only low leakage process (1.2V/1.5V) 2 or 64% under the same conditions. Since low operating power is often more important for infrastructure applications than low leakage power, high-performance processes are the best choice for this type of application.
Transistor design
The same process transistor can also have different switching threshold voltage (VT). For example, the switching speed of the low VT transistor is faster, the leakage current of the high VT transistor is smaller, and the wafer only needs to be used in the part that affects the speed. Low VT transistors, other circuits use high VT transistors to save power. The designer's component library should contain the basic logic gates (NAND, NOR, INVERT, etc.) made up of high VT and low VT transistors, and they sometimes use middle-VT transistors. In general, logic gates composed of high VT transistors should be used as much as possible in order to meet important performance requirements.
Component operating point: voltage and frequency
Several component clock supply methods save power:
â— multiple clock domain;
â— dynamic frequency scaling (dynamic frequency scaling);
â— Clock gating.
In addition to the clock, adjusting the voltage also reduces power consumption:
â— static voltage adjustment;
â— Dynamic voltage / frequency adjustment;
â— Multiple voltage domain.
Langrui Energy (Shenzhen) Co.,Ltd , https://www.langruibattery.com