• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2006년도 하계종합학술대회
• /
• pp.1027-1028
• /
• 2006

A case study in low-power PostPC Platform is presented. We introduce an S3C2460 Mobile SoC Processor and Implementation of Embedded Linux on out platform. This Processor is designed to Multimedia & Telecommunication Applications. We focuse on the verification of S3C2460 Processor and operation of Embedded Linux OS on it.

• 정보학연구
• /
• 제4권4호
• /
• pp.51-63
• /
• 2001

잉여수계(Residue Number System)는 각 모듈러스에 자리올림수의 전달이 필요 없고, 병렬 구조를 이루므로, 디지털 신호처리 및 신경망 처리기와 같은 전용 프로세서 설계에서 사용된다. 그러나, 크기 비교 및 부호 검출시에 혼합기수변환(Mixed Radix Conversion)이 요구되며, 이는 전체 연산 속도를 저해하는 요인이 된다. 그러므로 본 논문에서는 혼합기수 변환의 속도를 향상시키기 위하여 잉여수 분할 방법을 개선한 혼합기수변환기를 설계하였다. 설계된 변환기는 기존의 변환기에 비하여 연산기의 크기는 증가하지만 연산시간은 최대 2배가 향상되었다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2005년도 추계종합학술대회
• /
• pp.711-714
• /
• 2005

We propose floating point arithmetic units for geometry operation of mobile 3D graphic processor. The proposed arithmetic units conform to the single precision format of IEEE standard 754-1985 that is a standard of floating point arithmetic. The rounding algorithm applies the nearest toward zero form. The proposed adder/subtraction unit and multiplier have one clock cycle latency, and the inversion unit has three clock cycle latency. We estimate the required numbers of arithmetic operation for Viewing transformation. The first stage of geometry operation is composed with translation, rotation and scaling operation. The translation operation requires three addition and the rotation operation needs three addition and six multiplication. The scaling operation requires three multiplication. The viewing transformation is performed in 15 clock cycles. If the adder and the multiplier have their own in/out ports, the viewing transformation can be done in 9 clock cycles. The error margin of proposed arithmetic units is smaller than $10^{-5}$ that is the request in the OpenGL standard. The proposed arithmetic units carry out operations in 100MHz clock frequency.

• ETRI Journal
• /
• 제34권1호
• /
• pp.44-54
• /
• 2012

Because of the intrinsic lack of internal-system observability and controllability in highly integrated multicore processors, very restricted access is allowed for the debugging of erroneous chip behavior. Therefore, the building of an efficient debug function is an important consideration in the design of multicore processors. In this paper, we propose a flexible on-chip debug architecture that embeds a special logic supporting the debug functionality in the multicore processor. It is designed to support run-stop-type debug functions that can halt and control the execution of the multicore processor at breakpoint events and inspect the possible causes of any errors. The debug architecture consists of the following three functional components: the core debug support block, the multicore debug support block, and the debug interface and control block. By embedding this debug infrastructure, the embedded processor cores within the multicore processor can be debugged simultaneously as well as independently. The debug control is performed by employing a JTAG-based scanning operation. We apply this on-chip debug architecture to build a debugger for a prototype multicore processor and demonstrate the validity and scalability of our approach.

• 한국통신학회:학술대회논문집
• /
• 한국통신학회 1991년도 추계종합학술발표회논문집
• /
• pp.92-95
• /
• 1991

Constant false alarm rate(CFAR) processors are useful for detecting radar targets in background for which all parameters in the statistical distribution are not known and may be nonstationary. The well known "cell averging" (CA) CFAR processor is known to yield best performance in homogeneous case, but exhibits severe performance in the presence of an interfering target in the reference window or/and in the region of clutter edges. The "order statistics"(OS) CFAR processor is known to have a good performance above two nonhomogeneous cases. The modified OS-CFAR processor, known as "trimmed mean"(TM) CFAR processor performs somewhat better than the OS-CFAR processor by judiciously trimming the ordered samples. This paper proposes and analyzes the performance of a new CFAR processor called the "maximum trimmed mean"(MX-TM) CFAR processor combining the "greatest of"(GO) CFAR and TM-CFAR processors. The MAX operation is included to control false alarms at clutter edges. Our analyses show that the proposed CFAR processor has similar performance TM- and OS-CFAR processors in homogeneous case and in the precence of interfering targets, but can control the false rate in clutter edges. Simulation results are presented to demonstrate the qualitative effects of various CFAR processors in nonhomogeneous clutter environments.

• 한국정보처리학회논문지
• /
• 제6권5호
• /
• pp.1225-1240
• /
• 1999

This paper proposes PASC(PArtitioned SCHeduler) processor architecture that equips with a number of functional unit and an individual scheduler paris. Every scheduler of the PASC processor can determine whether a unit instruction can be issued to the associated functional unit or it is to be waited until next cycle caused by a resource collision or data dependencies. In the PASC processor, only the functional unit with a resource collision or data dependencies waits by executing a NOP(No OPeration) instruction and the other functional units execute their own instructions. Therefore we can expect the code compaction effect on the PASC processor. Thus, the last instruction of a loop at certain iteration and the very first instruction of the loop at the next iteration can be scheduled simultaneously if the two instructions do not incur any resource collision or data dependencies. Therefore, we can expect that such two instructions without any resource collision and data dependencies are packed into the same very long instruction word and thus, the two instructions are executed concurrently at run time. As a result, we can shorten execution cycles of a loop comparing to the execution of the loop on a traditional VLIW or SVLIW processor architecture. Simulation result also promises faster execution of loops on a PASC processor architecture than those on a VLIW and SVLIW processor architecture.

• 한국컴퓨터산업학회논문지
• /
• 제9권4호
• /
• pp.157-170
• /
• 2008

본 연구에서는 메사츄세츄공과대학 마이크로전자 연구소에서 개발된 프로세서 PARWAN(PAR-1)으로 불리우는 줄여진 프로세서(a reduced processor)를 VHDL을 이용하여 Behavioral Leve에서 기술하고 Dataflow Level에서 상호 연결하여 기술하였고, VHDL로 설계된 CPU의 동작을 확인하고 시뮬레이션하기 위하여 Test-bench 방식을 이용하였다. <중략> 제시된 방식은 설계의 정보교환이 용이하고 동작의 표현이 정확하고 간결하였으며, 설계의 문서화가 용이하며, 구성된 프로세서의 동작을 확인하기가 용이하였다. VHDL의 Behavioral 기술은 설계자에게 설계된 시스템을 확인할 때 많은 도움을 주었으며 Dataflow 기술은 설계의 버스연결과 레지스터 구조를 확인할 때 유용하게 사용할 수 있었다.

• Carbon letters
• /
• 제9권4호
• /
• pp.303-307
• /
• 2008

The studies on integrated operation of fuel cell with fuel processor are very essential prior to its commercialization. In this study, Polymer Electrolyte Membrane Fuel Cell (PEMFC) was operated with a fuel processor, which is mainly composed of two parts, methanol steam reforming reaction and preferential oxidation (PROX). In fuel processor, ICI 33-5 (CuO 50%, ZnO 33%, $Al_2O_3$ 8%, BET surface area: $66\;m^2g^{-1}$) catalyst and CuO-$CeO_2$ catalyst were used for methanol steam reforming, preferential oxidation (PROX) respectively. PEMFC was operated by hydrogen fuel generated from fuel processor. The resulting gas from PROX reactor is used to operate PEMFC equipped with our prepared anode and cathode catalyst. PtRu/C catalyst gives more tolerance to CO.

• Journal of Semiconductor Engineering
• /
• 제2권2호
• /
• pp.130-135
• /
• 2021

Recently, deep neural networks (DNNs) are actively used for action control so that an autonomous system, such as the robot, can perform human-like behaviors and operations. Unlike recognition tasks, the real-time operation is essential in action control, and it is too slow to use remote learning on a server communicating through a network. New learning techniques, such as reinforcement learning (RL), are needed to determine and select the correct robot behavior locally. In this paper, we propose an energy-efficient DNN processor with a LUT-based processing engine and near-zero skipper. A CNN-based facial emotion recognition and an RNN-based emotional dialogue generation model is integrated for natural HRI system and tested with the proposed processor. It supports 1b to 16b variable weight bit precision with and 57.6% and 28.5% lower energy consumption than conventional MAC arithmetic units for 1b and 16b weight precision. Also, the near-zero skipper reduces 36% of MAC operation and consumes 28% lower energy consumption for facial emotion recognition tasks. Implemented in 65nm CMOS process, the proposed processor occupies 1784×1784 um2 areas and dissipates 0.28 mW and 34.4 mW at 1fps and 30fps facial emotion recognition tasks.

• 한국정밀공학회지
• /
• 제17권5호
• /
• pp.116-123
• /
• 2000

This paper presents an NC code post-processor that adjusts feedrates to keep the variation of metal removal rate along the tool paths minimum. Metal removal rate is estimated by virtually machining the part, whose surface model is built from a series of NC codes defined in operation plan, with cutting-tool-assembly models, whose geometry are defined in a machining database. The NC code post-processor modifies the feedrates by the adjustment rules, which are based on the machining knowledge for effective machining. This paper illustrates a procedure fur grouping machining conditions and we also show how to determine an adjustment rule for a machining-condition group. An example part was machined and it shows that the variation of cutting force was dramatically reduced after applying the NC code post-processor. The NC code post-processor is expected to increase productivity while maintaining the quality of the machined part.