Microcontrollers (MCUs) go deep into people’s application lives, and MCUs can be seen in almost all devices. After the introduction of DSP digital signal processors and FPU floating-point arithmetic unit functions in MCUs, MCUs have greatly expanded the applicable range of components. In recent years, Many MCU manufacturers have launched a variety of integration solutions for their products, whether it is product strategy or market segmentation, it also makes the MCU market more diversified.
It is not easy to question the trend of MCU (Microcontroller Unit) going deep into daily life applications, especially when MCU is used for function optimization or market segmentation purposes, DSP (digital signal processor) digital signal processor or FPU (Floating Point Unit) floating point operation unit The function integration greatly expands the applicable field of MCU.
If FPU or DSP is used for the purpose of importing, generally adding FPU and DSP integration architecture to the MCU, the main purpose is to consider the design direction under cost, especially in the early semiconductor components, SOC (System on Chip) system single chip and MCU have a price If you only need SDP or FPU for computing acceleration, but don’t want to choose a high unit price SOC, then MCU products that integrate DSP or FPU hardware acceleration units can not only provide better operating efficiency, but also perform better in cost control. Excellent.
The cost of MCU integrated chip packaging has plummeted to increase MCU functions and expand application spaceFrom the perspective of early SOC products, hardware accelerators equipped with DSP and FPU are important features of SOC products. Among them, the application direction of DSP and FPU is mainly to accelerate the processing of audio and image processing, and the continuous optimization of process technology, the cost of SOC Gradually getting closer to the MCU, under the 32-bit or even 64-bit architecture, the MCU has also begun to have solutions that combine DSP or FPU hardware acceleration units.
Let’s take a look at the advantages of the MCU plus the hardware acceleration unit. Add FPU to the MCU. The most direct benefit is that the MCU is used to process similar FPU calculations in the early days. Due to the limitation of the MCU’s own calculation architecture, the calculation results will be relatively slow. Long, and when hardware accelerators are introduced to process floating-point operations, because hardware calls or data transfers can calculate data through hardware, the MCU's memory resources consumed in floating-point operations can be reduced by at least 10% due to hardware acceleration integration.
Of course, from the point of view of purpose, regardless of whether the MCU integrates the FPU hardware acceleration unit, the floating-point calculation needs to use the MCU's existing computing power to obtain results, but the premise is that the calculation process will consume more computing time and hardware resources For systems that can wait and do not need to provide immediate response, it is natural to not consider the MCU solution integrating FPU, but for integration requirements that require high system performance and feedback response speed, the benefit of MCU combined with FPU is not only optimized for computing resource consumption , Energy-saving advantages and other effects, on the contrary, it is the utility of accelerating system response and performance improvement. It is the most direct and important function improvement of MCU combined with FPU hardware acceleration, and it also allows MCU to respond to more complex integration tasks.
High-level numerical calculations use hardware acceleration to meet design needsIn the early MCU components, the 8-bit architecture was still the mainstream application direction. In the data processing and calculation processing of the MCU, there were inherent processing limitations due to the architecture problem. For example, when the MCU performs decimal point and fraction processing operations, because If the number of 4-bit or 8-bit elements is limited, finite values ​​must be used for processing. The limitation of numerical results is exchanged for the purpose of simplifying processing complexity and performance requirements. This error caused by numerical processing is called "cutoff error" , The truncation error will also expand the error value due to the limitation of using the MCU for data calculation.
When the MCU integrates FPU hardware acceleration, when computing the same type of data processing, such as IoT or terminal sensor applications, there are often data capture and processing requirements for converting external analog sensing data into digital data. At this time, the FPU/DSP hardware acceleration unit integrated by the MCU can not only process the sensing data more quickly and speed up the system response, but also introduce advanced calculations to reduce data calculation errors.
In practical applications, the FPU hardware accelerator itself cannot completely solve the problem of error expansion, so there will be application purposes under different hardware acceleration integration architectures such as FPU and DSP. For example, through the DSP hardware accelerator, special data types can be targeted Higher-speed and reliable arithmetic processing output, such as DSP can use instructions to perform a variety of operations, processing such as fast Fourier transform (fast Fourier transform; FFT) or finite impulse response (Finite impulse response; FIR) is important in advanced operations In addition, resource-consuming computing needs, even through a single cycle of instructions, can process single instruction multiple data (Single InstrucTIon MulTIple Data; SIMD) computing needs, and the MCU can also obtain advanced enhanced benefits in terms of advanced numerical processing.
Different hardware acceleration units of FPU/DSP have complementary effectsAlthough the integration of FPU or DSP is basically different in architecture and application direction, in fact, the two are respectively for data operation and signal processing corresponding to various algorithm applications. The functions of the two can be said to have complementary effects and it is difficult to be independent. Split. From the ARM Cortex-M4 perspective, if only the DSP hardware acceleration processor is provided, the FPU floating-point arithmetic accelerator is not set, which will cause application restrictions. Because in the Cortex-M4 application, if only the digital signal processing acceleration hardware supports, there will be less floating Point calculation support, if the development demand side encounters the need for numerical advanced calculation acceleration, it will cause design flexibility restrictions, or it will require external function chip support, or use the original calculation resources to respond to numerical advancement Computational requirements will limit the application possibilities of Cortex-M4 because of numerical processing performance.
The same situation also occurs in microcontroller applications with only FPU and no DSP. The application functions of DSP or FPU are complementary to each other. Independent integration does not produce synergy for the configuration of the microcontroller, but will become a development. Limitation of the path.
Furthermore, from the development direction of the new generation of IoT products, through the application direction of Sensor Fusion as an example, if the Sensor Fusion concept is to integrate multiple sensors in a single system to work together, the system requires high-level values ​​and signals Only with processing power can the key numerical signals be extracted from complex data.
As for sensor fusion, it can be combined with real-time adjustment, control and correction processing. The DSP and FPU co-processing can achieve high precision and high efficiency for precise analysis of captured data, especially the existing Sensor Fusion has already achieved gyro The sensor, accelerometer, temperature, pressure and even touch sensing are all made in the same module. The DSP and FPU must be used to pre-screen out the relatively precise and efficient signal acquisition and preprocessing of the sensing data to provide Relatively efficient systems have more efficient sensing value processing mechanisms.
DSP digital filtering application can improve the quality of sensing signal captureIn addition, another advantage of integrating FPU in MCU is that it can make good use of its computing characteristics in the system. For example, use digital algorithms for digital filtering applications to capture values, and further use hardware-accelerated digital algorithms for waveform processing of signals. Or data reprocessing to form a convenient way to improve the data-to-noise ratio (SNR). The digital filter can also use the calculation mechanism to optimize the filtering effect of different degrees. This is because the microcontroller is used to sense popular heart rate and blood content. Physiological information such as oxygen level and exercise value, or applications such as digital meters, smart meters, etc., can solve the data optimization effect of inverted signal distortion due to noise or environmental noise at the end data, and optimize the signal waveform signal quality obtained by the terminal. It is more conducive to subsequent processing or data use.
In order to optimize the end application, the integration of hardware acceleration units in microcontrollers is also a trend, not just DSP or FPU hardware acceleration units, for example, some microcontrollers have added VMU hardware acceleration units to the architecture to deal with the focus of motor applications. It is separated from the existing FPU floating-point arithmetic hardware acceleration function to accelerate the application performance of the overall microcontroller by means of coordination and division of labor.
What’s interesting is that, for different markets and computing requirements, the microcontroller differentiates between computing clocks and distinguishes different applications and market segments with the most practical computing performance, and integrates the hardware acceleration required for different applications. Units have also become an important demarcation of product market positioning. For example, microcontrollers for the wearable computing application market can only integrate FPU and DSP hardware to accelerate positioning market segments in terms of power consumption, sensor integration, and component cost. In the application of high-end microcontrollers, there are even solutions that directly integrate hardware graphics engines to directly meet the application requirements of industrial human-machine interface terminals. In addition, there are also a variety of different market needs such as automotive electronics and IoT. Hardware acceleration unit configuration combination, application architecture to meet different integration requirements.
Another purpose for microcontrollers to integrate DSP and FPU hardware acceleration units is to integrate hardware acceleration units without adopting external solutions to configure hardware acceleration computing requirements. Its biggest advantage lies in the ultimate cost optimization, because the electronic circuit The board can save more space on the carrier board, and the use of a single chip can improve the overall efficiency of computing. At the software development level, simple call and data transfer reprocessing can be used under an integrated architecture to meet the data computing output of application services. For efficiency requirements, even the finished product developed can use consistent debugging analysis tools to directly analyze and correct errors in the system, and the efficiency and speed of development and design can be improved.
High Quality Tablet Ram 4GB/8GB/16GB SODIMM UDIMM DDR4 8gb Laptop 2400mhz Ram
Model
DDR4 SODIMM
Brand
Semsotai /OEM /ODM /OBM
ECC
Non-ECC
Capacity
4GB/8GB/16GB
Chipsets
Samsung /SK Hynix/Mircon/Spec tek
Frequency
2133MHz/2400MHz/2666MHz/3200MHz
Speed
PC4-17000/PC4-19200/PC4-21300/PC4-25600
Componets Configuration
(256m/512m/1G) x8/x16
Memory CAS Lantency
CL15/CL17/CL19/CL21/CL22
Supply Voltage
1.2V
Pin Count
260PIN
PCB layer
8 /10 layer high speed PCB
Certificate
CE/FCC/ROHS
Operation temperature
0C-95C
Compability
100% Work with all motherboard
Ddr4 Laptop Memory,Ddr4 Rgb Ram,Ddr3 Ram 8Gb Pc 1600,Game Ddr3 8Gb
MICROBITS TECHNOLOGY LIMITED , https://www.hkmicrobits.com