Abstract: The present disclosure provides a memory control circuit configured to precede a data-reading process with a memory. For the data-reading process, the memory transmits a DQ and a DQS indicating a time to read the DQ. The DQS includes a preamble. The memory control circuit includes a control circuit and a sampling circuit. The control circuit is configured to generate an enabling signal. The sampling circuit coupled to the control circuit is configured to sample the DQS based on the enabling signal in order to determine a sampling level. The control circuit determines whether the sampling level matches a signal level of the preamble or not.

Abstract: An ODT circuit is connected to a memory module and includes a first transmission line, a first ODT, a second ODT, a first switch circuit, a third ODT, a fourth ODT, a second switch circuit, and an ODT control logic. The first and second ODTs are coupled to a first node on the first transmission line. The first switch circuit includes a first switch and a second switch, and is driven according to the first control signal. The third and the fourth ODTs are coupled to a second node on the first transmission line. The second switch circuit includes a third switch and a fourth switch, and is driven according to the second control signal. The ODT control logic outputs the first control signal and the second control signal to control the first switch circuit and the second switch circuit to be turned on at different timings.

Abstract: A memory signal phase difference calibration circuit includes: a clock generator providing clocks allowing a physical layer (PHY) circuit of DDR SDRAM to generate a data input/output signal (DQ) and a data strobe signal (DQS) for accessing a storage circuit; a calibration control circuit outputting a phase control signal according to an adjustment range to adjust the phase of a target signal (DQ or DQS), and outputting a calibration control signal; an access control circuit reading storage data representing predetermined data from the storage circuit according to the calibration control signal; a comparison circuit comparing the predetermined data with the storage data to output a result allowing the calibration control circuit to alter the adjustment range accordingly; and a phase controller outputting a clock control signal according to the phase control signal to set the phase of a target clock used for the PHY circuit generating the target signal.

Abstract: A memory test method is provided that includes the steps outlined below. The memory controller performs data-writing and data-reading on a memory module. When a quantity of read data is incorrect, a data-strobe enable signal is calibrated to perform data reading. When there is one of less than one piece of negative edge data reading content, a sampling unit is triggered. When the quantity of read data increases, the condition that the data-strobe signal is not received is determined. When the quantity does not increase, the memory controller is inspected. When there is more than one piece of read data, the burst mode setting of the memory module is inspected. When the quantity is correct and the content is not correct, a transmission circuit setting and the sampling unit are inspected. When the quantity and the content are correct, the test flow is terminated.

Abstract: A memory control device, which includes a signal generating circuit, a data writing circuit and a repeating circuit. The repeating circuit is coupled to the data writing circuit. The signal generating circuit is configured to generate a data strobe signal and send the data strobe signal to a memory. The data strobe signal comprises a preamble signal. The data writing circuit is configured to write a series of data to the memory according to the data strobe signal. The repeating circuit is configured to repeat a first data of the series of data in a period of the preamble signal.

Abstract: A memory test method is provided that includes the steps outlined below. The memory controller performs data-writing and data-reading on a memory module. When a quantity of read data is incorrect, a data-strobe enable signal is calibrated to perform data reading. When there is one of less than one piece of negative edge data reading content, a sampling unit is triggered. When the quantity of read data increases, the condition that the data-strobe signal is not received is determined. When the quantity does not increase, the memory controller is inspected. When there is more than one piece of read data, the burst mode setting of the memory module is inspected. When the quantity is correct and the content is not correct, a transmission circuit setting and the sampling unit are inspected. When the quantity and the content are correct, the test flow is terminated.

Abstract: A memory control device, which includes a signal generating circuit, a data writing circuit and a repeating circuit. The repeating circuit is coupled to the data writing circuit. The signal generating circuit is configured to generate a data strobe signal and send the data strobe signal to a memory. The data strobe signal comprises a preamble signal. The data writing circuit is configured to write a series of data to the memory according to the data strobe signal. The repeating circuit is configured to repeat a first data of the series of data in a period of the preamble signal.

Abstract: The present invention provides an overlay target. The overlay target includes a plurality of first pattern blocks and a plurality of second pattern blocks. The first pattern blocks and the second patterns blocks are arranged in array by being separated by at least one first gaps stretching along a first direction and at least one second gaps stretching along a second direction. Each first pattern block is composed of a plurality of first stripe patterns stretching along a third direction, and each second pattern block is composed of a plurality of second stripe patterns stretching along a fourth direction. The first direction is orthogonal to the second direction, the third direction and the fourth direction are 45 degrees relative to the first direction.

Abstract: A memory physical layer interface circuit electrically connected between a memory controller and a memory device is provided. The memory physical layer interface circuit includes a dock generation module and first-in-first-out (FIFO) modules. The clock generation module generates a reference clock signal and output related clock signals. The reference clock signal is transmitted to the memory device. Each of the FIFO modules writes the input information therein transmitted by the memory controller according to a write-related clock signal and retrieves the input information therefrom according to one of the output related clock signals to generate an output signal. The output signal is transmitted to the memory device to operate the memory device. The write-related clock signal is generated by dividing a frequency of one of the output related clock signals.

Abstract: An asymmetry compensation method used in a lithography overlay process and including steps of: providing a first substrate, wherein a circuit layout is disposed on the first substrate, a first mask layer and a second mask layer together having an x-axis allowable deviation range and an y-axis allowable deviation range relative to the circuit layout are stacked sequentially on the circuit layout, wherein the x-axis allowable deviation range is unequal to the y-axis allowable deviation range; and calculating an x-axis final compensation parameter and an y-axis final compensation parameter base on the unequal x-axis allowable deviation range and the y-axis allowable deviation range.

Abstract: A memory physical layer interface circuit electrically connected between a memory controller and a memory device is provided. The memory physical layer interface circuit includes a dock generation module and first-in-first-out (FIFO) modules. The clock generation module generates a reference clock signal and output related clock signals. The reference clock signal is transmitted to the memory device. Each of the FIFO modules writes the input information therein transmitted by the memory controller according to a write-related clock signal and retrieves the input information therefrom according to one of the output related clock signals to generate an output signal. The output signal is transmitted to the memory device to operate the memory device. The write-related clock signal is generated by dividing a frequency of one of the output related clock signals.

Abstract: An overlap mark set is provided to have at least a first and a second overlap marks both of which are located at the same pattern layer. The first overlap mark includes at least two sets of X-directional linear patterns, having a preset offset a1 therebetween; and at least two sets of Y-directional linear patterns, having the preset offset a1 therebetween. The second overlap mark includes at least two sets of X-directional linear patterns, having a preset offset b1 therebetween; and at least two sets of Y-directional linear patterns, having the preset offset b1 therebetween. The preset offsets a1 and b1 are not equal.

Abstract: The present invention provides a photo-mask for manufacturing structures on a semiconductor substrate, which comprises a photo-mask substrate, a first pattern, a second pattern and a forbidden pattern. A first active region, a second active region are defined on the photo-mask substrate, and a region other than the first active region and the second active region are defined as a forbidden region. The first pattern is disposed in the first active region and corresponds to a first structure on the semiconductor substrate. The second pattern is disposed in the second active region and corresponds to a second structure on the semiconductor substrate. The forbidden pattern is disposed in the forbidden region, wherein the forbidden pattern has a dimension beyond resolution capability of photolithography and is not used to form any corresponding structure on the semiconductor substrate. The present invention further provides a method of manufacturing semiconductor structures.

Abstract: A method of correcting an overlay error includes the following steps. First, an overlay mark disposed on a substrate is captured so as to generate overlay mark information. The overlay mark includes at least a pair of first mark patterns and at least a second mark pattern above the first mark patterns. Then, the overlay mark information is calculated to generate an offset value between two first mark patterns and to generate a shift value between the second mark pattern and one of the first mark patterns. Finally, the offset value is used to compensate the shift value so as to generate an amended shift value.

Abstract: A memory control circuit includes a comparator, an eye width measuring circuit and a calibration circuit, wherein the comparator is arranged to compare a data signal with a reference voltage to generate a compared data signal; the eye width measuring circuit is coupled to the comparator, and is arranged to measure an eye width of the compared data signal to generate a measuring result; and the calibration circuit is coupled to the comparator and the eye width measuring circuit, and is arranged to adjust a level of the reference voltage according to the measuring result.

Abstract: A double patterning method comprises the following steps. First of all, a target layer and a mask layer stacked thereon are provided. Next, a first pattern opening is formed in the mask layer, and a width of the first pattern opening is measured to obtain a measuring value. Then, a second pattern opening is formed in the mask layer based on the measuring value, wherein the second pattern opening and the first pattern opening are co-planar. Finally, a bias trimming process is performed to trim the first pattern opening and the second pattern opening.

Abstract: a memory control circuit includes a comparator, an eye width measuring circuit and a calibration circuit, wherein the comparator is arranged to compare a data signal with a reference voltage to generate a compared data signal; the eye width measuring circuit is coupled to the comparator, and is arranged to measure an eye width of the compared data signal to generate a measuring result; and the calibration circuit is coupled to the comparator and the eye width measuring circuit, and is arranged to adjust a level of the reference voltage according to the measuring result.

Abstract: The present invention provides a photo-mask for manufacturing structures on a semiconductor substrate, which comprises a photo-mask substrate, a first pattern, a second pattern and a forbidden pattern. A first active region, a second active region are defined on the photo-mask substrate, and a region other than the first active region and the second active region are defined as a forbidden region. The first pattern is disposed in the first active region and corresponds to a first structure on the semiconductor substrate. The second pattern is disposed in the second active region and corresponds to a second structure on the semiconductor substrate. The forbidden pattern is disposed in the forbidden region, wherein the forbidden pattern has a dimension beyond resolution capability of photolithography and is not used to form any corresponding structure on the semiconductor substrate. The present invention further provides a method of manufacturing semiconductor structures.

Abstract: An asymmetry compensation method used in a lithography overlay process and including steps of: providing a first substrate, wherein a circuit layout is disposed on the first substrate, a first mask layer and a second mask layer together having an x-axis allowable deviation range and an y-axis allowable deviation range relative to the circuit layout are stacked sequentially on the circuit layout, wherein the x-axis allowable deviation range is unequal to the y-axis allowable deviation range; and calculating an x-axis final compensation parameter and an y-axis final compensation parameter base on the unequal x-axis allowable deviation range and the y-axis allowable deviation range.

Abstract: A method of correcting an overlay error includes the following steps. First, an overlay mark disposed on a substrate is captured so as to generate overlay mark information. The overlay mark includes at least a pair of first mark patterns and at least a second mark pattern above the first mark patterns. Then, the overlay mark information is calculated to generate an offset value between two first mark patterns and to generate a shift value between the second mark pattern and one of the first mark patterns. Finally, the offset value is used to compensate the shift value so as to generate an amended shift value.