Abstract: Any of a plurality of contact plugs which reaches a diffusion layer serving as a drain layer of an MOS transistor has an end provided in contact with a lower surface of a thin insulating film provided selectively on an interlayer insulating film. A phase change film constituted by GST to be a chalcogenide compound based phase change material is provided on the thin insulating film, and an upper electrode is provided thereon. Any of the plurality of contact plugs which reaches the diffusion layer serving as a source layer has an end connected directly to an end of a contact plug penetrating an interlayer insulating film.

Abstract: A phase change memory includes a memory cell with a phase change element storing data according to level change of a resistance value in association with phase change, a write circuit converting the phase change element to an amorphous state or a polycrystalline state according to the logic of write data in a write operation mode, a read circuit reading out stored data from the phase change element in a readout operation mode, and a discharge circuit applying a discharge voltage to the phase change element to remove electrons trapped in the phase change element in a discharge operation mode. Accordingly, variation in the resistance value at the phase change element can be suppressed.

Abstract: Any of a plurality of contact plugs which reaches a diffusion layer serving as a drain layer of an MOS transistor has an end provided in contact with a lower surface of a thin insulating film provided selectively on an interlayer insulating film. A phase change film constituted by GST to be a chalcogenide compound based phase change material is provided on the thin insulating film, and an upper electrode is provided thereon. Any of the plurality of contact plugs which reaches the diffusion layer serving as a source layer has an end connected directly to an end of a contact plug penetrating an interlayer insulating film.

Abstract: Until the number of pulse application n reaches 12, as a first-half pulse, a pulse is set to have a width fixed to 2 ms, and its voltage is increased every time. As a latter-half pulse, the pulse is set to have a width fixed to 3 ms and the pulse voltage is increased every time until the maximum voltage is attained. After the maximum voltage is attained, first, the pulse of a width of 3 ms is applied twice, the pulse of a width of 4 ms with the maximum voltage is applied twice, and the pulse of a width of 5 ms with the maximum voltage is applied twice. Even after the maximum voltage is attained, change over time of a threshold voltage can be more linear. Thus, a non-volatile semiconductor memory device allowing efficient programming operation and erasing operation in a short period of time can be provided.

Abstract: Until the number of pulse application n reaches 12, as a first-half pulse, a pulse is set to have a width fixed to 2 ms, and its voltage is increased every time. As a latter-half pulse, the pulse is set to have a width fixed to 3 ms and the pulse voltage is increased every time until the maximum voltage is attained. After the maximum voltage is attained, first, the pulse of a width of 3 ms is applied twice, the pulse of a width of 4 ms with the maximum voltage is applied twice, and the pulse of a width of 5 ms with the maximum voltage is applied twice. Even after the maximum voltage is attained, change over time of a threshold voltage can be more linear. Thus, a non-volatile semiconductor memory device allowing efficient programming operation and erasing operation in a short period of time can be provided.

Abstract: Until the number of pulse application n reaches 12, as a first-half pulse, a pulse is set to have a width fixed to 2 ms, and its voltage is increased every time. As a latter-half pulse, the pulse is set to have a width fixed to 3 ms and the pulse voltage is increased every time until the maximum voltage is attained. After the maximum voltage is attained, first, the pulse of a width of 3 ms is applied twice, the pulse of a width of 4 ms with the maximum voltage is applied twice, and the pulse of a width of 5 ms with the maximum voltage is applied twice. Even after the maximum voltage is attained, change over time of a threshold voltage can be more linear. Thus, a non-volatile semiconductor memory device allowing efficient programming operation and erasing operation in a short period of time can be provided.

Abstract: A dummy cell having a low threshold voltage is disposed in a memory cell array in alignment with a memory cell. A dummy cell with a low threshold voltage adjacent to a selected memory cell column is selected, and a source-side local bit line of the selected memory cell is coupled to a global bit line via such dummy cell. Since the source-side local bit line is coupled to a ground node at its both ends, source resistance of the memory cell can be reduced, and dependency of the source resistance of the memory cell on the position within the memory cell array can also be reduced. This allows for reducing dependency of source resistance of a memory cell on the position within the memory cell array and on the temperature in a nonvolatile semiconductor memory device.

Abstract: Until the number of pulse application n reaches 12, as a first-half pulse, a pulse is set to have a width fixed to 2 ms, and its voltage is increased every time. As a latter-half pulse, the pulse is set to have a width fixed to 3 ms and the pulse voltage is increased every time until the maximum voltage is attained. After the maximum voltage is attained, first, the pulse of a width of 3 ms is applied twice, the pulse of a width of 4 ms with the maximum voltage is applied twice, and the pulse of a width of 5 ms with the maximum voltage is applied twice. Even after the maximum voltage is attained, change over time of a threshold voltage can be more linear. Thus, a non-volatile semiconductor memory device allowing efficient programming operation and erasing operation in a short period of time can be provided.

Abstract: A dummy cell having a low threshold voltage is disposed in a memory cell array in alignment with a memory cell. A dummy cell with a low threshold voltage adjacent to a selected memory cell column is selected, and a source-side local bit line of the selected memory cell is coupled to a global bit line via such dummy cell. Since the source-side local bit line is coupled to a ground node at its both ends, source resistance of the memory cell can be reduced, and dependency of the source resistance of the memory cell on the position within the memory cell array can also be reduced. This allows for reducing dependency of source resistance of a memory cell on the position within the memory cell array and on the temperature in a nonvolatile semiconductor memory device.

Abstract: In a flash memory, a threshold voltage of a memory transistor is decreased quickly by increasing a rising speed of a pulse voltage of an erasing pulse signal train during the first period of an erasing operation. In response to the threshold voltage of the memory transistor becoming lower than a threshold voltage of a reference transistor, the threshold voltage of the memory transistor is decreased slowly by reducing the rising speed of the pulse voltage of the erasing pulse signal train. Therefore, the erasing time can be reduced and depletion of the memory transistor can be prevented.

Abstract: In a flash memory, a threshold voltage of a memory transistor is decreased quickly by increasing a rising speed of a pulse voltage of an erasing pulse signal train during the first period of an erasing operation. In response to the threshold voltage of the memory transistor becoming lower than a threshold voltage of a reference transistor, the threshold voltage of the memory transistor is decreased slowly by reducing the rising speed of the pulse voltage of the erasing pulse signal train. Therefore, the erasing time can be reduced and depletion of the memory transistor can be prevented.

Abstract: Until the number of pulse application n reaches 12, as a first-half pulse, a pulse is set to have a width fixed to 2 ms, and its voltage is increased every time. As a latter-half pulse, the pulse is set to have a width fixed to 3 ms and the pulse voltage is increased every time until the maximum voltage is attained. After the maximum voltage is attained, first, the pulse of a width of 3 ms is applied twice, the pulse of a width of 4 ms with the maximum voltage is applied twice, and the pulse of a width of 5 ms with the maximum voltage is applied twice. Even after the maximum voltage is attained, change over time of a threshold voltage can be more linear. Thus, a non-volatile semiconductor memory device allowing efficient programming operation and erasing operation in a short period of time can be provided.