摘要:
An entry including multiple bits of unit cells each storing data bit is coupled to a match line. The match line is supplied with a charging current having a restricted current value smaller than a match line current flowing in a one-bit miss state in one entry, but larger than a match line current flowing in an all-bit match state in one entry. A precharge voltage level of a match line is restricted to a voltage level of half a power supply voltage or smaller. Power consumption in a search cycle of a content addressable memory can be reduced, and a search operation speed can be increased.
摘要:
An entry including multiple bits of unit cells each storing data bit is coupled to a match line. The match line is supplied with a charging current having a restricted current value smaller than a match line current flowing in a one-bit miss state in one entry, but larger than a match line current flowing in an all-bit match state in one entry. A precharge voltage level of a match line is restricted to a voltage level of half a power supply voltage or smaller. Power consumption in a search cycle of a content addressable memory can be reduced, and a search operation speed can be increased.
摘要:
An entry including multiple bits of unit cells each storing data bit is coupled to a match line. The match line is supplied with a charging current having a restricted current value smaller than a match line current flowing in a one-bit miss state in one entry, but larger than a match line current flowing in an all-bit match state in one entry. A precharge voltage level of a match line is restricted to a voltage level of half a power supply voltage or smaller. Power consumption in a search cycle of a content addressable memory can be reduced, and a search operation speed can be increased.
摘要:
An entry including multiple bits of unit cells each storing data bit is coupled to a match line. The match line is supplied with a charging current having a restricted current value smaller than a match line current flowing in a one-bit miss state in one entry, but larger than a match line current flowing in an all-bit match state in one entry. A precharge voltage level of a match line is restricted to a voltage level of half a power supply voltage or smaller. Power consumption in a search cycle of a content addressable memory can be reduced, and a search operation speed can be increased.
摘要:
An entry including multiple bits of unit cells each storing data bit is coupled to a match line. The match line is supplied with a charging current having a restricted current value smaller than a match line current flowing in a one-bit miss state in one entry, but larger than a match line current flowing in an all-bit match state in one entry. A precharge voltage level of a match line is restricted to a voltage level of half a power supply voltage or smaller. Power consumption in a search cycle of a content addressable memory can be reduced, and a search operation speed can be increased.
摘要:
Read data line pairs, write data line pairs, a spare read data line pair, and a spare write data line pair are provided extending in the column direction over a memory cell array. Spare bit repair is performed by replacing a data line pair. Column redundancy control circuit changes the timing for outputting the result of spare determination for a data write mode and for a data read mode. A semiconductor memory device suitable for merging with a logic and capable of reducing the current consumption and achieving a higher operation frequency is provided.
摘要:
The object of the present invention is to obtain a communication device which improves latency and throughput in data transfer processing by reducing the number of accesses to a local bus. A bus interface 1a receives an user data from an AALLSI 4, and, when a host bus cannot be occupied, temporarily stores the user data in a data receiving buffer area 11. The capacity of the data receiving buffer area 11 in the bus interface 1a is set so as not to be full during normal data processing.
摘要:
A data transfer controller (14) decides that both of a host bus (100) and a local bus (101) have been acquired by a bus interface LSI 1b by the fact that a host bus use permission GNTH from a host bus arbiter (2b) and a local bus use permission GNTL from a local bus arbiter (4) are both activated, in which case, the data transfer controller (14) controls a host bus DMA controller (12) and a local bus DMA controller (13) to make selectors SE1 and SE2 establish a connection between the host bus (100) and the local bus (101) through an inner data path (11b) having no buffer. This configuration improves data transfer rate through the host bus (100) and the local bus (101).
摘要:
Incoming lines (I0 to I7) are connected to a space switch (2) through input data latches (1). The space switch (2) is connected to a normal/test changeover switch (12), which is connected to a normal/test changeover switch (13) through serial-to-parallel converting circuits (3), common buffer memories (4) and parallel-to-serial converting circuits (5). Space switches (6) are connected to the normal/test changeover switch (13). Outgoing lines (O0 to O7) are connected to the space switches 6 through output data latches (8). Connection states in the switches (2, 6) are placed in transposed relation to each other by a transposed connection generating circuit (10) in a test operation, so that the switches (2, 6) are directly connected to each other through the switches (12, 13). Predetermined data applied to the incoming lines are intactly used as expected values for judgement of the normal or abnormal operation of the set of switches of matrix structure.
摘要:
An object is to provide a method for preventing the occurrence of variations in time resolution by providing a calibration process to a TDC at the time of start up and further preventing the increase in circuit scale by reducing the redundancy of delay elements. A calibration of a multiphase oscillator TDC and a vernier TDC is carried out at the time of power-on. In the calibration, a timing input to be input to the vernier TDC is selected from output signals of DCCO based on a reference clock. Also, data is defined as an output signal which is adjacent to the output signal of DCCO mentioned above and proceeds in phase, and the delay therebetween is derived. By repeating it to all of the output signals, the one cycle of the output signal of DCCO is derived.