CN103970709A - Communication method for FFT coprocessor and main processor - Google Patents
Communication method for FFT coprocessor and main processor Download PDFInfo
- Publication number
- CN103970709A CN103970709A CN201410177438.7A CN201410177438A CN103970709A CN 103970709 A CN103970709 A CN 103970709A CN 201410177438 A CN201410177438 A CN 201410177438A CN 103970709 A CN103970709 A CN 103970709A
- Authority
- CN
- China
- Prior art keywords
- fft
- bus
- cpu
- data
- coprocessor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Advance Control (AREA)
Abstract
本发明提供一种FFT协处理器与主处理器通信方法,该方法包括如下步骤:将CPU和DMA控制器均连接到AHB总线;DMA控制器集成AHB总线到APB总线的桥;FFT协处理器集成APB总线的从控制器;FFT协处理器发出中断请求;CPU响应所述中断请求;DMA控制器将FFT的运算结果搬移到系统内存;通信结束。通过本发明提供的方法,解决了慢速FFT协处理器和主处理器数据交换的问题;根据FFT协处理器所需要传输的数据的特点采用DMA进行FFT协处理器与主处理器的数据交换,提高了CPU和总线的利用率。
The invention provides a communication method between an FFT coprocessor and a main processor, the method comprising the steps of: connecting both the CPU and the DMA controller to the AHB bus; the DMA controller integrating the bridge from the AHB bus to the APB bus; and the FFT coprocessor The slave controller integrating the APB bus; the FFT coprocessor sends an interrupt request; the CPU responds to the interrupt request; the DMA controller moves the FFT operation result to the system memory; the communication ends. Through the method provided by the invention, the problem of data exchange between the slow FFT coprocessor and the main processor is solved; according to the characteristics of the data that the FFT coprocessor needs to transmit, DMA is used to exchange data between the FFT coprocessor and the main processor , Improved CPU and bus utilization.
Description
技术领域technical field
本发明涉及集成电路领域通信方法,具体涉及一种FFT协处理器与主处理器通信方法。The invention relates to a communication method in the field of integrated circuits, in particular to a communication method between an FFT coprocessor and a main processor.
背景技术Background technique
傅里叶变换是对信号进行频域分析得到信号频域特性的最常用的分析方法。离散傅里叶变换的提出给数字信号的频域分析提供了理论基础。然而直接进行离散傅里叶变换,运算量极高,在快速傅里叶变换(FFT)被提出之前应用并不广泛。FFT的提出极大的缩减了傅里叶变换的运算量,让傅里叶变换应用于实时系统成为可能,此时信号的傅里叶分析才开始广泛应用起来。Fourier transform is the most commonly used analysis method for frequency domain analysis of signals to obtain signal frequency domain characteristics. The proposal of discrete Fourier transform provides a theoretical basis for frequency domain analysis of digital signals. However, the direct discrete Fourier transform has a very high amount of calculation, and it was not widely used before the fast Fourier transform (FFT) was proposed. The introduction of FFT greatly reduces the computational load of Fourier transform, making it possible for Fourier transform to be applied to real-time systems. At this time, Fourier analysis of signals began to be widely used.
FFT可以通过通用DSP,FPGA和专用ASIC实现。专用普通的处理器也能够通过软件配置实现FFT的功能,但由于普通处理器并不像DSP一样有专门针对复杂运算优化的总线结构和运算单元,使用普通处理器进FFT运算时,FFT大量的乘法和加法运算会极大的占用CPU资源,导致功很高而效率很低。所以一般不会使用通用的CPU来进行FFT运算。随着集成电路的发展,单个芯片上集成多个功能模块成为一个发展趋势,这种芯片被称作SOC(System On Chip)。对于特定的应用,FFT也可以被设计成单独的专用模块和CPU集成到一块芯片上作为该CPU的协处理器辅助CPU实习FFT的功能。而CPU只负责简单的控制,就能够得到FFT的分析结果。从而CPU可以从大量的运算中解脱出来去做擅长的控制工作,以达到提高效率,降低功耗的目的。FFT can be realized by general-purpose DSP, FPGA and special-purpose ASIC. Dedicated ordinary processors can also realize the FFT function through software configuration, but because ordinary processors do not have bus structures and arithmetic units optimized for complex operations like DSP, when using ordinary processors for FFT operations, a large number of FFT Multiplication and addition operations will greatly occupy CPU resources, resulting in high work and low efficiency. Therefore, general-purpose CPUs are generally not used for FFT operations. With the development of integrated circuits, it has become a development trend to integrate multiple functional modules on a single chip. This chip is called SOC (System On Chip). For specific applications, FFT can also be designed as a separate dedicated module and CPU integrated into a chip as a coprocessor of the CPU to assist the CPU to practice the function of FFT. The CPU is only responsible for simple control, and the analysis results of FFT can be obtained. In this way, the CPU can be freed from a large number of calculations to do the control work that it is good at, so as to achieve the purpose of improving efficiency and reducing power consumption.
FFT协处理器中一般含有一块嵌入式的RAM(Random-Access Memory)用于暂存FFT的中间运算结果和最终运算结果。由于FFT协处理器实时处理ADC送进来的数据,最终的运算结果不能够长久保存,必须在下一次的运算结果输入进来之前搬移到系统内存。The FFT coprocessor generally contains an embedded RAM (Random-Access Memory) for temporarily storing the intermediate and final calculation results of the FFT. Since the FFT coprocessor processes the data sent by the ADC in real time, the final calculation result cannot be stored for a long time and must be moved to the system memory before the next calculation result is input.
发明内容Contents of the invention
为了克服上述现有技术的不足,本发明提供一种FFT协处理器通过DMA和CPU进行通信的方法,实现FFT协处理器和CPU的高效数据通信。In order to overcome the deficiencies of the above-mentioned prior art, the present invention provides a method for FFT coprocessor to communicate with CPU through DMA, so as to realize efficient data communication between FFT coprocessor and CPU.
为了实现上述发明目的,本发明采取如下技术方案:In order to realize the above-mentioned purpose of the invention, the present invention takes the following technical solutions:
一种FFT协处理器与主处理器通信方法,其特征在于,所述方法包括如下步骤:A kind of FFT coprocessor and main processor communication method, it is characterized in that, described method comprises the steps:
A.将CPU和DMA控制器均连接到AHB总线;A. Connect both the CPU and the DMA controller to the AHB bus;
B.DMA控制器集成AHB总线到APB总线的桥;B. The DMA controller integrates the bridge from the AHB bus to the APB bus;
C.FFT协处理器集成APB总线的从控制器;C. FFT coprocessor integrates the slave controller of APB bus;
D.FFT协处理器发出中断请求;D. The FFT coprocessor sends an interrupt request;
E.CPU响应所述中断请求;E. CPU responds to the interrupt request;
F.DMA控制器将FFT的运算结果搬移到系统内存;F. The DMA controller moves the FFT operation result to the system memory;
G.通信结束。G. Communication ends.
优选地,步骤A包括:通过选择器控制所述AHB总线的控制权;于DMA控制器上设置AHB总线的主控制器和从控制器,主控制器和从控制器从AHB总线上接收数据,并在获得总线控制权后发送数据到所述AHB总线。Preferably, step A includes: controlling the control right of the AHB bus through a selector; setting a master controller and a slave controller of the AHB bus on the DMA controller, and the master controller and the slave controller receive data from the AHB bus, And send data to the AHB bus after obtaining the bus control right.
优选地,步骤B中,AHB总线到APB总线的桥完成AHB数据到APB数据的协议转换功能。Preferably, in step B, the bridge from the AHB bus to the APB bus completes the protocol conversion function from AHB data to APB data.
优选地,步骤C中,所述APB总线的从控制器与外界交换数据;DMA上的APB控制器连接FFT协处理器,实现基于APB协议的数据通路。Preferably, in step C, the slave controller of the APB bus exchanges data with the outside; the APB controller on the DMA is connected to the FFT coprocessor to realize a data path based on the APB protocol.
优选地,步骤D中,所述FFT协处理器发出中断请求包括:当FFT协处理器有数据运算完成时,FFT协处理器向CPU发出中断请求,请求CPU搬移本次FFT运算得到的数据。Preferably, in step D, the FFT coprocessor sending an interrupt request includes: when the FFT coprocessor completes data operations, the FFT coprocessor sends an interrupt request to the CPU, requesting the CPU to move the data obtained by this FFT operation.
优选地,步骤E中,CPU响应所述中断请求包括:CPU根据中断向量调用中断处理程序;中断处理程序将需要搬移的数据的起始地址、结束地址和搬移模式通过AHB总线配置到DMA控制器中的寄存器中。Preferably, in step E, the CPU responding to the interrupt request includes: the CPU calls the interrupt handler according to the interrupt vector; the interrupt handler configures the start address, end address and transfer mode of the data to be moved to the DMA controller through the AHB bus in the register.
优选地,步骤F包括:CPU完成配置后释放AHB总线的控制权,DMA控制器获得AHB总线的控制权,并根据CPU的配置将FFT对应地址内存储的FFT的运算结果搬移到系统内存中。Preferably, step F includes: the CPU releases the control right of the AHB bus after completing the configuration, and the DMA controller obtains the control right of the AHB bus, and moves the operation result of the FFT stored in the address corresponding to the FFT to the system memory according to the configuration of the CPU.
优选地,步骤G包括:DMA控制器完成数据搬移工作后,发出中断;CPU接收中断得知工作完成,通信过程结束。Preferably, step G includes: after the DMA controller completes the data transfer work, it sends an interrupt; the CPU receives the interrupt and knows that the work is completed, and the communication process ends.
与现有技术相比,本发明的有益效果在于:Compared with prior art, the beneficial effect of the present invention is:
本发明的FFT协处理器是针对电网信号进行FFT分析的专用的模块,可以直接分析模数转换器(ADC)将外界模拟信号转换得到的数字信号;将该信号进行FFT分析之后通过DMA(Direct Memory Access)控制器传送到系统内存中;The FFT coprocessor of the present invention is a dedicated module for FFT analysis of power grid signals, and can directly analyze the digital signal obtained by converting an external analog signal by an analog-to-digital converter (ADC); Memory Access) controller is sent to the system memory;
本专利解决了慢速FFT协处理器和主处理器数据交换的问题;This patent solves the problem of data exchange between the slow FFT coprocessor and the main processor;
本专利根据FFT协处理器所需要传输的数据的特点采用DMA进行FFT协处理器与主处理器的数据交换。通过使用DMA控制器,CPU将DMA进行简单的配置即可实现将FFT产生的数据直接搬移到内存当中,降低了CPU的数据搬移负载,提高了CPU和总线的利用率。According to the characteristics of the data to be transmitted by the FFT coprocessor, this patent uses DMA to exchange data between the FFT coprocessor and the main processor. By using the DMA controller, the CPU can simply configure the DMA to move the data generated by FFT directly to the memory, which reduces the data moving load of the CPU and improves the utilization of the CPU and the bus.
附图说明Description of drawings
图1是本发明实施例中通信系统结构图。Fig. 1 is a structural diagram of a communication system in an embodiment of the present invention.
具体实施方式Detailed ways
下面结合附图对本发明作进一步详细说明。The present invention will be described in further detail below in conjunction with the accompanying drawings.
本发明提供的方法包括如下步骤:Method provided by the invention comprises the steps:
1、CPU和DMA控制器均连接到高速的AHB总线上,通过更高一级的选择器来控制谁对总线有控制权。DMA控制器上有AHB总线的主控制器和从控制器,可以从AHB总线上接收数据也可以获得总线控制权后发送数据到AHB总线。1. Both the CPU and the DMA controller are connected to the high-speed AHB bus, and a higher-level selector is used to control who has control over the bus. The DMA controller has a master controller and a slave controller of the AHB bus, which can receive data from the AHB bus or obtain bus control and send data to the AHB bus.
2、DMA上直接集成AHB总线到APB总线的桥,完成AHB数据到APB数据的协议转换功能。2. The DMA directly integrates the bridge from the AHB bus to the APB bus to complete the protocol conversion function from AHB data to APB data.
3、FFT协处理器集成APB的从控制器以实现与外界的数据交换。DMA上的APB控制器直接连接到FFT协处理器上,实现一条专用的基于APB协议的数据通路,实现FFT协处理器与DMA的直接通信。3. The FFT coprocessor integrates the slave controller of APB to realize data exchange with the outside world. The APB controller on the DMA is directly connected to the FFT coprocessor to realize a dedicated data path based on the APB protocol, and realize the direct communication between the FFT coprocessor and the DMA.
4、当FFT有数据运算完成时向主处理器发出中断请求,请求主处理器来搬移本次FFT运算得到的数据。4. Send an interrupt request to the main processor when the FFT data operation is completed, requesting the main processor to move the data obtained by this FFT operation.
5、主处理器响应FFT模块的中断请求,根据相应的中断向量调用中断处理程序。中断处理程序将需要搬移的数据的起始地址,结束地址,搬移模式通过AHB总线配置到DMA控制器中的相关寄存器中。5. The main processor responds to the interrupt request of the FFT module, and calls the interrupt handler according to the corresponding interrupt vector. The interrupt handler configures the start address, end address, and transfer mode of the data to be moved to the relevant registers in the DMA controller through the AHB bus.
6、CPU完成配置后释放去做其他工作,DMA控制器获得AHB总线的控制权,根据CPU的配置将FFT对应地址内存储的FFT的运算结果搬移到系统内存中。6. After the CPU completes the configuration, it is released to do other work. The DMA controller obtains the control right of the AHB bus, and moves the FFT operation result stored in the FFT corresponding address to the system memory according to the CPU configuration.
7、DMA完成数据搬移工作后,发出中断,CPU接收中断得知工作完成。实施例一如下:7. After the DMA completes the data moving work, it sends an interrupt, and the CPU receives the interrupt to know that the work is completed. Embodiment one is as follows:
电网频率很低(50Hz或者60Hz),对电网进行谐波分析一般需要到50到100次谐波左右。若要得到较高的频域分辨率,和较高的精度,FFT输出的信号所需要的存储空间约为10-20k。例如,当电网频率为50Hz时,要求能够分析到100次谐波,则信号带宽至少为5000,为方便实现,取每周波采样256个点,则奈奎斯特采样率为12.8k。一般采用插值滤波FFT的方法来在较低的FFT的分析的点数上达到较高的分析精度。本应用做4096点FFT分析,输出信号为复数,位宽取24比特,则一共需要的存储空间为4k*2*24/8=24kB。因此,每次FFT分析将会产生24kB的数据。The frequency of the power grid is very low (50Hz or 60Hz), and the harmonic analysis of the power grid generally requires about 50 to 100 harmonics. To obtain higher frequency domain resolution and higher precision, the storage space required by the FFT output signal is about 10-20k. For example, when the power grid frequency is 50Hz, it is required to be able to analyze the 100th harmonic, and the signal bandwidth is at least 5000. For the convenience of implementation, 256 sampling points per cycle are taken, and the Nyquist sampling rate is 12.8k. Generally, the method of interpolation filtering FFT is used to achieve higher analysis accuracy at a lower number of FFT analysis points. This application does 4096-point FFT analysis, the output signal is a complex number, and the bit width is 24 bits, so the total storage space required is 4k*2*24/8=24kB. Therefore, each FFT analysis will generate 24kB of data.
参考图1,当FFT协处理器完成一次FFT运算时,数据全部暂存在FFT协处理器的一块输出RAM上。这时,FFT协处理器向CPU提起中断请求。CPU收到中断请求,找到对应的中断向量,跳到对应的中断地址,执行相应中断处理程序。该程序控制处理器通过AHB总线将配置数据写入到DMA中的寄存器中。其中配置数据包括待搬移的数据的原始地址和长度,待搬移数据的目标地址的起始地址,搬移方式。Referring to Fig. 1, when the FFT coprocessor completes an FFT operation, all the data is temporarily stored on an output RAM of the FFT coprocessor. At this time, the FFT coprocessor raises an interrupt request to the CPU. The CPU receives the interrupt request, finds the corresponding interrupt vector, jumps to the corresponding interrupt address, and executes the corresponding interrupt handler. The program controls the processor to write the configuration data into the registers in the DMA through the AHB bus. The configuration data includes the original address and length of the data to be moved, the starting address of the target address of the data to be moved, and the way of moving.
处理器将DMA中的寄存器配置完成后,释放AHB总线的控制权,去完成其他工作。此时DMA控制器按照CPU的配置获得AHB总线的控制权。同时,DMA控制器通过与FFT协处理器相连的专用APB总线按照APB协议的规定读取FFT协处理器中的数据。读到的数据经过DMA控制器的协议转换后通过AHB总线写入到系统内存中。数据完成传送后DMA向处理器提起中断。After the processor configures the registers in the DMA, it releases the control of the AHB bus to complete other tasks. At this time, the DMA controller obtains the control right of the AHB bus according to the configuration of the CPU. At the same time, the DMA controller reads the data in the FFT coprocessor according to the provisions of the APB protocol through the dedicated APB bus connected with the FFT coprocessor. The read data is written into the system memory through the AHB bus after the protocol conversion of the DMA controller. After the data transfer is completed, the DMA raises an interrupt to the processor.
最后应当说明的是:以上实施例仅用以说明本发明的技术方案而非对其限制,尽管参照上述实施例对本发明进行了详细的说明,所属领域的普通技术人员应当理解:依然可以对本发明的具体实施方式进行修改或者等同替换,而未脱离本发明精神和范围的任何修改或者等同替换,其均应涵盖在本发明的权利要求范围当中。Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the present invention can still be Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall be covered by the scope of the claims of the present invention.
Claims (8)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410177438.7A CN103970709A (en) | 2014-04-29 | 2014-04-29 | Communication method for FFT coprocessor and main processor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410177438.7A CN103970709A (en) | 2014-04-29 | 2014-04-29 | Communication method for FFT coprocessor and main processor |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103970709A true CN103970709A (en) | 2014-08-06 |
Family
ID=51240230
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410177438.7A Pending CN103970709A (en) | 2014-04-29 | 2014-04-29 | Communication method for FFT coprocessor and main processor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103970709A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104750518A (en) * | 2015-03-10 | 2015-07-01 | 昂纳信息技术(深圳)有限公司 | Upgrade control method and control module |
CN105403769A (en) * | 2015-09-30 | 2016-03-16 | 中国农业大学 | Circuit structure based on FFT short-time Fourier analysis and control method thereof |
CN106959936A (en) * | 2016-01-08 | 2017-07-18 | 福州瑞芯微电子股份有限公司 | A kind of the hardware-accelerated of FFT realizes device and method |
CN107291642A (en) * | 2016-04-11 | 2017-10-24 | 罗伯特·博世有限公司 | Microcontroller, control device and motor vehicles |
CN107644003A (en) * | 2017-09-29 | 2018-01-30 | 北京智芯微电子科技有限公司 | A kind of computing Co-processor Module of main control chip |
CN108021454A (en) * | 2017-12-28 | 2018-05-11 | 努比亚技术有限公司 | A kind of method, terminal and the computer-readable storage medium of processor load equilibrium |
CN114880259A (en) * | 2022-07-12 | 2022-08-09 | 北京象帝先计算技术有限公司 | Data processing method, device, system, electronic device and storage medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7065629B2 (en) * | 2002-11-18 | 2006-06-20 | Rf Micro Devices, Inc. | Address translation logic for use in a GPS receiver |
US7071859B1 (en) * | 2005-06-30 | 2006-07-04 | Sigmatel, Inc. | System and method for scheduling access to an analog-to-digital converter and a microprocessor |
CN201041677Y (en) * | 2007-02-05 | 2008-03-26 | 凤凰微电子(中国)有限公司 | Intelligent card supporting high-performance computing, large-capacity storage and high-speed transmission |
CN101159729A (en) * | 2007-09-05 | 2008-04-09 | 杭州华三通信技术有限公司 | Method and device for fast processing packet |
CN101324839A (en) * | 2007-06-15 | 2008-12-17 | 松下电器产业株式会社 | information processing device |
CN101901200A (en) * | 2010-08-11 | 2010-12-01 | 烽火通信科技股份有限公司 | Method for realizing double advanced high-performance bus (AHB) Master interface-based on-chip direct memory access (DMA) controller |
CN102231142A (en) * | 2011-07-21 | 2011-11-02 | 浙江大学 | Multi-channel direct memory access (DMA) controller with arbitrator |
-
2014
- 2014-04-29 CN CN201410177438.7A patent/CN103970709A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7065629B2 (en) * | 2002-11-18 | 2006-06-20 | Rf Micro Devices, Inc. | Address translation logic for use in a GPS receiver |
US7071859B1 (en) * | 2005-06-30 | 2006-07-04 | Sigmatel, Inc. | System and method for scheduling access to an analog-to-digital converter and a microprocessor |
CN201041677Y (en) * | 2007-02-05 | 2008-03-26 | 凤凰微电子(中国)有限公司 | Intelligent card supporting high-performance computing, large-capacity storage and high-speed transmission |
CN101324839A (en) * | 2007-06-15 | 2008-12-17 | 松下电器产业株式会社 | information processing device |
CN101159729A (en) * | 2007-09-05 | 2008-04-09 | 杭州华三通信技术有限公司 | Method and device for fast processing packet |
CN101901200A (en) * | 2010-08-11 | 2010-12-01 | 烽火通信科技股份有限公司 | Method for realizing double advanced high-performance bus (AHB) Master interface-based on-chip direct memory access (DMA) controller |
CN102231142A (en) * | 2011-07-21 | 2011-11-02 | 浙江大学 | Multi-channel direct memory access (DMA) controller with arbitrator |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104750518A (en) * | 2015-03-10 | 2015-07-01 | 昂纳信息技术(深圳)有限公司 | Upgrade control method and control module |
CN105403769A (en) * | 2015-09-30 | 2016-03-16 | 中国农业大学 | Circuit structure based on FFT short-time Fourier analysis and control method thereof |
CN105403769B (en) * | 2015-09-30 | 2018-05-11 | 中国农业大学 | A kind of circuit structure and its control method based on FFT Short Time Fourier Analysis |
CN106959936A (en) * | 2016-01-08 | 2017-07-18 | 福州瑞芯微电子股份有限公司 | A kind of the hardware-accelerated of FFT realizes device and method |
CN107291642A (en) * | 2016-04-11 | 2017-10-24 | 罗伯特·博世有限公司 | Microcontroller, control device and motor vehicles |
CN107644003A (en) * | 2017-09-29 | 2018-01-30 | 北京智芯微电子科技有限公司 | A kind of computing Co-processor Module of main control chip |
CN108021454A (en) * | 2017-12-28 | 2018-05-11 | 努比亚技术有限公司 | A kind of method, terminal and the computer-readable storage medium of processor load equilibrium |
CN114880259A (en) * | 2022-07-12 | 2022-08-09 | 北京象帝先计算技术有限公司 | Data processing method, device, system, electronic device and storage medium |
CN114880259B (en) * | 2022-07-12 | 2022-09-16 | 北京象帝先计算技术有限公司 | Data processing method, device, system, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103970709A (en) | Communication method for FFT coprocessor and main processor | |
WO2020001459A1 (en) | Data processing method, remote direct memory access network card, and device | |
CN105045763B (en) | A kind of PD Radar Signal Processing Systems and its Parallel Implementation method based on FPGA+ multi-core DSPs | |
CN110196824B (en) | Method and device for realizing data transmission and electronic equipment | |
CN103345461B (en) | Based on the polycaryon processor network-on-a-chip with accelerator of FPGA | |
CN109447241B (en) | A Dynamic Reconfigurable Convolutional Neural Network Accelerator Architecture for the Internet of Things | |
CN106775599A (en) | Multi-computing-unit coarse-grained reconfigurable system and method for recurrent neural network | |
CN105630735A (en) | Coprocessor based on reconfigurable computational array | |
CN104820657A (en) | Inter-core communication method and parallel programming model based on embedded heterogeneous multi-core processor | |
CN103984560A (en) | Embedded reconfigurable system based on large-scale coarseness and processing method thereof | |
CN104317770A (en) | Data storage structure and data access method for multiple core processing system | |
CN103902498B (en) | A kind of software definition server system towards Heterogeneous Computing and method | |
CN115098412B (en) | Peripheral access controller, data access device and corresponding method, medium and chip | |
CN103995797A (en) | Method for communication between FFT coprocessor and main processor | |
CN104915213A (en) | Partial reconfiguration controller of reconfigurable system | |
Ramagond et al. | A review and analysis of communication logic between PL and PS in ZYNQ AP SoC | |
CN110108938A (en) | A kind of real-time electromagnetic spectrum analysis system and method based on GPU data flow driven | |
CN109726800B (en) | Operation method, device and related product | |
CN102708079B (en) | Be applied to the method and system of the control data transmission of microcontroller | |
CN107678781B (en) | Processor and method for executing instructions on processor | |
CN105511919B (en) | A kind of the dynamic local reconfigurable controller and its control method of reconfigurable system | |
CN102880762A (en) | Multi-core processor platform based on Noc (Network operating Center) framework and design method thereof | |
CN106873447A (en) | A kind of robot model's algorithm implementation method based on FPGA | |
CN203746056U (en) | Multi-operand four-arithmetic mixed fixed-point arithmetic operation controller | |
CN103217681B (en) | Tree-shaped topological mechanism multiprocessor sonar signal processing method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C41 | Transfer of patent application or patent right or utility model | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20160325 Address after: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant after: State Grid Corporation of China Applicant after: State Grid Smart Grid Institute Applicant after: Electric Power Research Institute of State Grid Shandong Electric Power Company Address before: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant before: State Grid Corporation of China Applicant before: State Grid Smart Grid Institute |
|
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140806 |