Abstract: A processor including a coarse grained array including a plurality of function units and a plurality of register files, wherein a loop to be executed by the coarse grained array is split into a plurality of sub-loops, and when an interrupt request occurs while executing the sub-loop in the coarse grained array, the interrupt request is processed after the executing of the sub-loop is completed.

Abstract: Provided are a method and apparatus for avoiding bank conflict. A first instruction that is one of access instructions that are predicted to cause the bank conflict is replaced with a second instruction by changing an execute timing of the first instruction to a timing prior to the execute timing of the first instruction so as for the access instructions not to cause the bank conflict. Next, a load/store unit that is scheduled to access the bank according to the first instruction accesses the bank and reads out a data from the bank at an execute timing of the second instruction, and after that, the load/store unit is allowed to be inputted the read data at the execute timing of the first instruction. Accordingly, although the access instructions that are predicted to cause the bank conflict are allocated to the load/store units, the bank conflict can be prevented, so that it is possible to avoid deterioration in performance due the occurrence of the bank conflict.

Abstract: Disclosed are a semiconductor device and a related method of manufacture. The semiconductor device comprises a semiconductor substrate, a conductive structure including contact regions and gate structures formed on the semiconductor substrate, a protection layer formed on the gate structures, an insulation layer formed on the protection layer, and a plurality of contacts directly contacting the contact regions and the semiconductor substrate through the insulation layer, wherein the contacts have substantially different heights from each other.

Abstract: The present invention provides nitrogenous compositions for forming a silicon nitride layer, wherein the nitrogenous composition comprises a hydrazine compound, an amine compound or a mixture thereof. The present invention further provides source compositions for forming a silicon nitride layer, wherein the source composition comprises a nitrogenous composition comprising a hydrazine compound, an amine compound or a mixture thereof, and a silicon source comprising hexachlorodisilane. Methods for forming silicon nitride layers are further provided. The silicon nitride layers provided herein may be formed on a substrate at a low temperature and may further exhibit improved breakdown voltage and an enhanced etch resistance.

Abstract: An apparatus and a method are provided for a parallel processing very long instruction word (VLIW) computer. The apparatus includes: an index code generation unit sequentially generating an index code, which is associated with a number of no operation (NOP) instruction word between effective instruction words, with respect to each of instruction word groups to be executed in a VLIW computer; an instruction compression unit sequentially deleting the NOP instruction word which corresponds to the index code with respect to each of instruction word groups; and an instruction word conversion unit converting the effective instruction words to include the index code, the effective instruction words corresponding to the NOP instruction words.

Abstract: A loop coalescing method and a loop coalescing device are disclosed. The loop coalescing method comprises removing an inner-most loop from among nested loops, so that an outer operation provided outside of the inner-most loop is performed when a condition of a conditional statement is satisfied, generating a guard code by applying an if-conversion method to the conditional statement, and converting a guard by using an instruction calculating the guard of the guard code, the instruction calculating the guard using a register where information related to a period of time corresponding to the number of iterations of the inner-most loop is stored.

Abstract: A memory access method includes: obtaining a, b, and c from a program code for accessing a memory with a triple loop in a program, a being a number of values which an inner-most loop variable of the triple loop may have, b being a number of values which a middle loop variable of the triple loop may have, and c being a number of values which an outer-most loop variable of the triple loop may have; obtaining a starting address of the memory accessed by the triple loop; and obtaining an a×b×c number of addresses of the memory accessed by the triple loop using the starting address and a function.

Abstract: Provided is a processor and method of performing speculative load instructions of the processor in which a load instruction is performed only in the case where the load instruction substantially accesses a memory. A load instruction for canceling operations is performed in other cases except the above case, so that problems occurring by accessing an input/output (I/O) mapped memory area and the like at the time of performing speculative load instructions can be prevented using only a software-like method, thereby improving the performance of a processor.

Abstract: A profiler which provides information to optimize an application specific architecture processor and a program for the processor is provided.

Abstract: A method and system for an early Z test in a tile-based three-dimensional rendering is provided. In the method and system for an early Z test, a portion which is not displayed to a user is removed prior to performing a rasterization process, and thereby performing the 3D rendering efficiently. The method includes segmenting a scene into tiles for performing a rendering with respect to a triangle; selecting a first tile of the tiles, which has a tile Z value less than a minimum Z value of the triangle; and performing the rendering with respect to the triangle in remaining tiles excluding the selected first tile of the tiles.

Abstract: An operation system and method of processing a user-defined extended operation are provided. The method includes using a software pipelining technology by enabling a processor to process a user-defined extended operation. An operation process system includes a plurality of functional units which are operable to process a primitive operation and a processor which is operable to process an extended operation according to a control of each of the functional units.

Abstract: A tile binning method including: dividing a scene for rendering a triangle, into a plurality of tiles; determining identification values of tile nodes of each of the tiles; and identifying a tile including an entirety or a part of the triangle from the tiles, based on the identification value of the tile nodes for each of the tiles.

Abstract: A method, medium, and apparatus to effectively handle an interrupt in a reconfigurable array. In the method, the reconfigurable array pauses execution of an operation when an interrupt request occurs, and after storing register values of a register to be used for handling the interrupt request, an interrupt service is performed by select processing units of the reconfigurable array in response to the interrupt request. Upon completion of the interrupt service, the register values are restored, and the reconfigurable array resumes execution of the operation.

Abstract: An apparatus and a method for processing an array in a loop in a computer system, including: applying loop unrolling to a multi-dimensional array included in a loop based on a predetermined unrolling factor to generate a plurality of unrolled multi-dimensional arrays; and transforming each of the plurality of unrolled multi-dimensional arrays into a one-dimensional array having an array subscript expression in a form of an affine function with respect to a loop counter variable.

Abstract: A processor including a coarse grained array including a plurality of function units and a plurality of register files, wherein a loop to be executed by the coarse grained array is split into a plurality of sub-loops, and when an interrupt request occurs while executing the sub-loop in the coarse grained array, the interrupt request is processed after the executing of the sub-loop is completed.

Abstract: A capacitor of a semiconductor device includes a cylinder type capacitor lower electrode, a dielectric layer, and an upper electrode. The upper electrode includes a metallic layer on the dielectric layer and a doped polySi1-xGex layer stacked on the metallic layer. Methods of forming these capacitors also are provided.

Abstract: In a method of manufacturing a ferroelectric capacitor, a lower electrode layer is formed on a substrate. The lower electrode layer includes at least one lower electrode film. A ferroelectric layer is formed on the lower electrode layer, and then an upper electrode layer is formed on the ferroelectric layer. A hard mask structure is formed on the upper electrode layer. The hard mask structure includes a first hard mask and a second hard mask. An upper electrode, a ferroelectric layer pattern and a lower electrode are formed by partially etching the upper electrode layer, the ferroelectric layer and the lower electrode layer using the hard mask structure. The hard mask structure may prevent damage to the ferroelectric layer and may enlarge an effective area of the ferroelectric capacitor so that the ferroelectric capacitor may have enhanced electrical and ferroelectric characteristics.

Abstract: Disclosed are a semiconductor device and a related method of manufacture. The semiconductor device comprises a semiconductor substrate, a conductive structure including contact regions and gate structures formed on the semiconductor substrate, a protection layer formed on the gate structures, an insulation layer formed on the protection layer, and a plurality of contacts directly contacting the contact regions and the semiconductor substrate through the insulation layer, wherein the contacts have substantially different heights from each other.

Abstract: A semiconductor device includes a capacitor having a bottom electrode, a dielectric layer formed on the bottom electrode, a top electrode formed on the dielectric layer, and a contact plug having a metal that is connected with the top electrode, wherein the top electrode includes a doped poly-Si1-xGex layer and a doped polysilicon layer epitaxially deposited on the doped poly-Si1-xGex layer and the contact plug makes a contact with the doped polysilicon layer.

Abstract: A method for forming a low-k dielectric layer for a semiconductor device using an ALD process including (a) forming predetermined interconnection patterns on a semiconductor substrate, (b) supplying a first and a second reactive material to a chamber having the substrate therein, thereby adsorbing the first and second reactive materials on a surface of the substrate, (c) supplying a first gas to the chamber to purge the first and second reactive materials that remain unreacted, (d) supplying a third reactive material to the chamber, thereby causing a reaction between the first and second materials and the third reactive material to form a monolayer, (e) supplying a second gas to the chamber to purge the third reactive material that remains unreacted in the chamber and a byproduct; and (f) repeating (b) through (e) a predetermined number of times to form a SiBN ternary layer having a predetermined thickness on the substrate.