• The KIPS Transactions:PartA
• /
• v.14A no.2
• /
• pp.107-116
• /
• 2007

In this paper, we present a method on the construction of multiple-valued circuits using Reed-Muller Expansions(RME). First, we discussed the input output interconnection of multiple valued function using Perfect Shuffle techniques and Kronecker product and designed the basic cells of performing the transform matrix and the reverse transform matrix of multiple valued RME using addition circuit and multiplication circuit of GF(4). Using these basic cells and the input-output interconnection technique based on Perfect Shuffle and Kronecker product, we implemented the multiple valued logic circuit based on RME. The proposed design method of multiple valued RME is simple and very efficient to reduce addition circuits and multiplication circuits as compared with other methods for same function because of using matrix transform based on modular structures. The proposed design method of multiple valued logic circuits is simple and regular for wire routing and possess the properties of concurrency and modularity of array.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.3
• /
• pp.613-623
• /
• 2011

In this paper, we present a method on the construction of quinternary logic circuits using Perfect shuffle. First, we discussed the input-output interconnection of quinternary logic function using Perfect Shuffle techniques and Kronecker product, and designed the basic cells of performing the transform matrix and the reverse transform matrix of quinternary Reed-Muller expansions(QRME) using addition circuit and multiplication circuit of GF(5). Using these basic cells and the input-output interconnection technique based on Perfect Shuffle and Kronecker product, we implemented the quinternary logic circuit based on QRME. The proposed design method of QRME is simple and very efficient to reduce addition circuits and multiplication circuits as compared with other methods for same logic function because of using matrix transform based on modular structures. The proposed design method of quinternary logic circuits is simple and regular for wire routing and possess the properties of concurrency and modularity of array.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.7
• /
• pp.956-962
• /
• 2020

Recently, as the size of data used in an embedded system increases, the need for an ECC decoding operation to robustly receive a massive data is emphasized. In this paper, we propose a method to accelerate the execution of computations that derive syndrome vectors when ECC decoding is performed using Hamming code in an embedded system with a built-in GPU. The proposed acceleration method uses the matrix-vector multiplication of the decoding operation using the CSR format, one of the data structures representing sparse matrix, and is performed in parallel in the CUDA kernel of the GPU. We evaluated the proposed method using a target embedded board with a GPU, and the result shows that the execution time is reduced when ECC decoding operation accelerated based on the GPU than used only CPU.

• Journal of IKEEE
• /
• v.28 no.1
• /
• pp.53-64
• /
• 2024

With the advancement of NLP(Natural Language Processing) models, conversational AI such as ChatGPT is becoming increasingly popular. To enhance processing speed and reduce power consumption, it is important to implement the Transformer algorithm, which forms the basis of the latest natural language processing models, in hardware. In particular, the multi-head attention and feed-forward network, which analyze the relationships between different words in a sentence through matrix multiplication, are the most computationally intensive core algorithms in the Transformer. In this paper, we propose a new variable systolic array based on the number of input words to enhance matrix multiplication speed. Quantization maintains Transformer accuracy, boosting memory efficiency and speed. For evaluation purposes, this paper verifies the clock cycles required in multi-head attention and feed-forward network and compares the performance with other multipliers.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.40 no.4
• /
• pp.51-57
• /
• 2003

The development of the communication network and the other network method can generate serious social problems. So, it is highly required to control security of network. These problems related security will be developed and keep up to confront with anti-security field such as hacking, cracking. The way to preserve security from hacker or cracker without developing new cryptographic algorithm is keeping the state of anti-cryptanalysis in a prescribed time by means of extending key-length. In this paper, we proposed M3 algorithm for the reduced processing time in the montgomery multiplication part. Proposed M3 algorithm using the matrix function M(.) and lookup table perform optionally montgomery multiplication with repeated operation. In this result, modified repeated operation part produce 30% processing rate than existed montgomery multiplicator. The proposed montgomery multiplication structured unit array method in carry generated part and variable length multiplication for eliminating bottle neck effect with the RSA cryptosystem. Therefore, this proposed montgomery multiplier enforce the real time processing and prevent outer cracking.

• Journal of Korea Foundry Society
• /
• v.28 no.6
• /
• pp.261-267
• /
• 2008

In-situ quasicrystalline icosahedral (I) phase reinforced Ti-based bulk metallic glass (BMG) matrix composites have been successfully fabricated by using two distinct thermal histories for BMG forming alloy. The BMG composite containing micron-scale Iphase has been introduced by controlling cooling rate during solidification, whereas nano-scale I-phase reinforced BMG composite has been produced by partial crystallization of BMG. For mechanical properties, micron-scale I-phase distributed BMG composite exhibited lower strength and plasticity compared to the monolithic BMG. On the other hand, nano-scale icosahedral phase embedded BMG composite showed enhanced strength and plasticity. These improved mechanical properties were attributed to the multiplication of shear bands and blocking of the shear band propagation in terms of isolation and homogeneous distribution of nanosize icosahdral phases in the glassy matrix, followed by stabilizing the mechanical and deformation instabilities.

• Journal of the Optical Society of Korea
• /
• v.6 no.3
• /
• pp.121-127
• /
• 2002

The operating principal of a ring laser gyroscope depends on the phase difference for the counter-propagating waves within a closed path. The reflecting mirrors mounted on the monoblock form the traveling waves. The manufacturing accuracy of the monoblock influences the traveling path of ray, the sensitivity of laser resonator for misalignments, and diffraction losses. A 3 $\times$ 3 ray transfer matrix was derived for optical components with centering and squaring errors in a ring resonator. The matrix can be utilized to predict the optical ray paths on the basis of the manufacturing errors of the monoblock as well as the misalignment of mirrors. Then the distance and orientation (o. slope) at the arbitrary plane inside the resonator along the ideal optical path can be calculated from the chain multiplication of the ray transfer matrix for each optical component in one round trip. We also show that the counter-propagating rays In a ring resonator with errors does not coincide in each round trip, which results in gain difference between two beams, and how these errors can be adjusted through the alignment procedure. Finally this 3 $\times$ 3 ray matrix formalism can be used to calculate the beam size and its displacement from the optical axis and the deviation at the diaphragm.

• Bulletin of the Korean Mathematical Society
• /
• v.38 no.4
• /
• pp.727-733
• /
• 2001

Let A be a commutative ring with nonzero identity and let M be an A-module. In this note we show that if $x = x_1, ..., x_n\; and\; y = y_1, ..., y_n$ both M-cosequence such that $Hx^T = y^T\; for\; some\; n\times n$ lower triangular matrix H over A, then the map $\beta_H : \;Ann_M(y_1,..., y_n)\;\rightarrow Ann_M(x_1,..., x_n)$ induced by multiplication by ｜H｜ is surjective.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1992.10a
• /
• pp.223-227
• /
• 1992

산업의 발전으로 정밀가공기술 및 측정기술 및 측정기술의 개발에 대한 요구가 급증하고 있으며, 특히 가공전 공작물의 장착정도, 팔레트 교환 위치정도와 가공후 가공정도 및 공작기계 자체의 정도 판별은 생산품의 정밀도와 직접 관계되는 중요한 요인이기 때문에 정확히 규명할 필요가 있다. 본 연구에서는 공작물의 장착오차와 기계오차가 존재하는 경우 Denavit-Hartenberg 접근법에 의하여 레퍼런스에 대한 공간상 linkage의 기하학적 표현을 Matrix Multiplication을 사용하여 Cutting Tool 과 Workpiece에 대한 공간상의 관계를 규명하였으며, 가공에 미치는 체적오차를 규명하고자 한다.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.32A no.9
• /
• pp.135-141
• /
• 1995

An optical digital parallel processor using Self-Pumped Phase Conjugate Mirror and liquid crystal spatial light modulator is presented and experimentally implemented. To use self-pumped PCM as memory, the mechanism for phase conjugation in two coupled interaction regions with the photorefractive crystal BaTiO$_{3}$ is investigated, especially the temporal behavior and effects of incident beam position. The optical design and implementation of matrix-vector multiplication using LCSLM and PCM memory is presented.