Abstract:
The present invention is provided to lessen load on a bus in the case of storing image data captured by a plurality of cameras into a semiconductor memory. To a semiconductor integrated circuit, a plurality of cameras and a semiconductor memory can be coupled. The semiconductor integrated circuit includes a plurality of first interfaces, a second interface, a bus, and a plurality of image processing modules. The image processing modules include a process of performing distortion correction on image data in a pre-designated region, and writing the image data in the region subjected to the distortion correction into the semiconductor memory via the bus and the second interface. By excluding image data out of the pre-designated region from an object of distortion correction in the image processing modules, the amount of image data transferred to the semiconductor memory is reduced.
Abstract:
A built-in memory is divided into the following two types: first memories 5 and 7 and second memories 4 and 6, and made accessible in parallel by third buses XAB and XDB and second buses YAB and YDB respectively. Thereby, a CPU core 2 can simultaneously transfer two data values from the built-in memory to a DSP engine 3. Moreover, the third buses XAB and XDB and the second buses YAB and YDB are also separate from first buses IAB and IDB to be externally interfaced and the CPU core 2 can access an external memory in parallel with the access to the second memories 4 and 6 and the first memories 5 and 7.
Abstract:
In microcomputers and digital signal processors in which a central processing unit for controlling the entire system and a digital signal processing unit having a product sum function required to process digital signals efficiently are mounted on one and the same chip, an increase in the number of processing steps caused by differing types of data handled by the calculators is prevented, thereby enhancing the efficiency of the digital signal processing.
Abstract:
An ATM cell processing apparatus including a DRAM for a frame producing buffer of a frame producing unit. In order to absorb the anisotropy of the access rate of the DRAM access, the random access mode of the DRAM access is always used. To compensate a drop in access rate in this case, the DRAM is arranged is an array form and each cell is divided. Resultant partial cell data are written into and read from respective DRAM banks in order. As a result, a fast cell buffer having a large capacity can be formed. The present cell buffer can be applied to a FIFO and the like as well.
Abstract:
A digital peak and valley detector including a peak value address register, a valley value address register, a peak value data register, a valley value data register, a peak comparator for comparing the value stored in the peak value data register with value data contained in the digital signal applied to the digital peak and valley detector and for causing the greater of the two to be stored in the peak value data register while simultaneously storing the address in the peak value address register and a valley comparator for comparing the valley value contained in the valley value data register with the value data contained in the digital signal applied to the digital peak and valley detector for determining which is less and for causing the value and the address of the smaller of the two to be stored respectively in the valley value data and the valley value address registers.
Abstract:
This invention relates to an information processing apparatus such as a digital signal processor and is applied particularly suitably to a digital filter.A plurality of data from initial value data till final value data relating to filtering coefficients of a digital filter are stored in a data memory, and are sequentially read out by an increment operation of an address arithmetic unit.A data arithmetic unit executes sequentially product and/or sum operations of a plurality of data that are sequentially read out and digital input signals that are sequentially inputted, to perform digital signal processing.The information processing apparatus is equipped particularly with means, which when an access address starts from an initial value, exceeds a final value and reaches a return address due to the increment operation, returns automatically the access address to the initial value. Therefore, a plurality of data stored in the data memory can be utilized repeatedly.Contrivances are made in order to set the number of a plurality of data that are stored in the data memory for repetition of use, to an arbitrary value.
Abstract:
A built-in memory is divided into the following two types: first memories 5 and 7 and second memories 4 and 6, and made accessible in parallel by third buses XAB and XDB and second buses YAB and YDB respectively. Thereby, a CPU core 2 can simultaneously transfer two data values from the built-in memory to a DSP engine 3. Moreover, the third buses XAB and XDB and the second buses YAB and YDB are also separate from first buses IAB and IDB to be externally interfaced and the CPU core 2 can access an external memory in parallel with the access to the second memories 4 and 6 and the first memories 5 and 7.
Abstract:
This data processor can satisfy both requests of a fast transition from a low power consumption state to an operating state and low power consumption, and a data processor has a program running state, a standby mode, a light standby mode, and a sleep mode. In the sleep mode, the supply of a synchronizing clock signal to a central processing unit (CPU) is stopped and the synchronizing clock signal is supplied to other circuit modules. In the standby mode, the frequency multiplication and frequency operation of a clock pulse generator are suspended and the supply of the synchronizing clock signal to the CPU and other circuit modules is stopped. In the light standby mode, the frequency multiplication and frequency division operation of the clock pulse generator are enabled and the supply of the synchronizing clock signal to the CPU and other circuit modules is stopped. In the light standby mode, the transition of the CPU to an instruction executable state is faster than in the standby mode and the lower power consumption than in the sleep mode is obtained.
Abstract:
In order to simplify the instruction prefetch architecture for use with the programs having few loops, and having instructions almost in linear and sequential addresses, the bus controller in accordance with the present invention for controlling the bus in an external memory includes a plurality of instruction buffers, flags each specific to each of instruction buffers, and a buffer controller circuit. The buffer controller circuit may allocate one of specific values that plural lower bits of an instruction address may take to each of the instruction buffers, and prefetch instructions to the instruction buffers each corresponding to a respective addresses designated to by the plural lower bits, from the address following a predetermined fetch address. The constitution for instruction prefetch as above may be implemented in a simpler manner than the controlling structure using address tags of a cache memory or the controlling structure using read-write pointer based on the counter in FIFO buffers.