• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2011.05a
• /
• pp.712-713
• /
• 2011

In this paper, the AMBA AXI slave unit that can verify the pipelined arithmetic unit is proposed and the 2-stage 16-bit pipelined multiplier is introduced as design example. The proposed AXI slave unit consists of input buffer block memory, control registers, pipelined arithmetic unit, control unit, output buffer block memory, and AXI slave interface unit. The main operational procedures are divided into the following steps, such as burst-mode input data loading for the input buffer memory, programming of control registers, arithmetic operations for block data in the input buffer memory, and burst-mode output data unloading from output buffer memory to host processor. Because the proposed AXI slave unit is general structure, it can be efficiently applicable to AMBA AXI and AHB slave unit with pipelined arithmetic unit.

• Proceedings of the KSR Conference
• /
• 2010.06a
• /
• pp.1068-1074
• /
• 2010

In high speed railway, event recorder is essential system for analyzing the cause of train accident. It stores train operation sent by train control system in safe memory unit. Crash protected memory, the safe memory unit for event recorder, keeps the stored contents from severe environment. For crash protected memory, we have designed the architecture of concrete enclosure and controller board. Proposed system provides large volume of memory capacity and fault tolerance architecture. For checking the characteristics of proposed crash protected memory specification, the simulation is executed. Simulation results shows the designed crash protected memory meets all requirements.

• Journal of the Semiconductor & Display Technology
• /
• v.22 no.2
• /
• pp.6-10
• /
• 2023

High-potential data can be predicted and stored in the cache to prevent cache misses, thus reducing the processor's request and wait times. As a result, the processor can work non-stop, hiding memory latency. By utilizing the temporal/spatial locality of memory access, the prefetcher introduced to improve the performance of these computers predicts the following memory address will be accessed. We propose a prefetcher that applies the GRU model, which is advantageous for handling time series data. Display the currently accessed address in binary and use it as training data to train the Gated Recurrent Unit model based on the difference (delta) between consecutive memory accesses. Finally, using a GRU model with learned memory access patterns, the proposed data prefetcher predicts the memory address to be accessed next. We have compared the model with the multi-layer perceptron, but our prefetcher showed better results than the Multi-Layer Perceptron.

• Proceedings of the IEEK Conference
• /
• 2005.11a
• /
• pp.825-828
• /
• 2005

In this paper, we propose a new nonvolatile unit memory cell and proposal peripheral circuit using the polymer material. Memory that relies on bistable behavior- having tow states associated with different resistances at the same applied voltage - has attracted much interest because of its nonvolatile properties. Such memory may also have other merits, including simplicity of structure and manufacturing, and the small size of memory cells. We have plotted the load line graphs for the use of a polymer memory character, hence we have designed in the band-gap reference shape of a write/erase drive, and then designed in the 2-stage differential amplifier shape of a sense amplifier in the consideration of a low current characteristic of a polymer memory cell. The simulation result shows that is has high gain about 80dB by sensing the very small current.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.17 no.3
• /
• pp.473-482
• /
• 2017

Hand gesture recognition is regarded as new Human Computer Interaction (HCI) technologies for the next generation of mobile devices. Previous hand gesture implementation requires a large memory and computation power for hand segmentation, which fails to give real-time interaction with mobile devices to users. Therefore, in this paper, we presents a low latency and memory-efficient hand segmentation architecture for natural hand gesture recognition. To obtain both high memory-efficiency and low latency, we propose a streaming hand contour tracing unit and a fast contour filling unit. As a result, it achieves 7.14 ms latency with only 34.8 KB on-chip memory, which are 1.65 times less latency and 1.68 times less on-chip memory, respectively, compare to the best-in-class.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.6
• /
• pp.1241-1245
• /
• 2009

As FFT(Fast Fourier Transform) processor is used in OFDM(Orthogonal Frequency Division Multiplesing) system. According to increase requirement about mobility and broadband, Research about low power and low area FFT processor is needed. So research concern in reduction of memory size and complex multiplier is in progress. Increasing points of FFT increase memory area of FFT processor. Specially, SRU(Signal Reordering Unit) has the most memory in FFT processor. In this paper, we propose a reduced method of memory size of SRU in FFT processor. SRU of 64, 1024 point FFT processor performed implementation by VerilogHDL coding and it verified by simulation. We select the APEX20KE family EP20k1000EPC672-3 device of Altera Corps. SRU implementation is performed by synthesis of Quartus Tool. The bits of data size decide by 24bits that is 12bits from real, imaginary number respectively. It is shown that, the proposed SRU of 64point and 1024point achieve more than 28%, 24% area reduction respectively.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2006.06a
• /
• pp.26-27
• /
• 2006

To develop tera-bit level SONOS flash memories, SONOS unit memory and 64 bit flash arrays are fabricated. The unit cells have both channel length and width of 30nm. The NAND & NOR arrays are fabricated on SOI wafer and patterned by E-beam. The unit cells represent good write/erase characteristics and reliability characteristics. SSL-NOR array have normal write/erase operation. These researches are leading the realization of Tera-bit level non-volatile nano flash memory.

• Proceedings of the KIEE Conference
• /
• 1988.07a
• /
• pp.567-570
• /
• 1988

This paper describes the design of FIR filtering DSP-chip that can be operated without programming. The proposed DSP-chip has not only the improvement of execution time but also selectivity of filter length from N=1 to N=128. Hilbert Transformer can be designed from this chip. FIR filter system is composed of Data memory/Control Unit, external memory and multiplier-accumulator. Data memory/Control Unit is laid out in this paper.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.3 no.3
• /
• pp.158-166
• /
• 2008

These days, mobile embedded systems adopt flash memory capable of XIP feature since they can reduce memory usage, power consumption, and software load time. XIP provides direct access to ROM and flash memory for processors. However, using XIP incurs unnecessary degradation of applications' performance because direct access to ROM and flash memory shows more delay than that to main memory. In this paper, we propose a memory management framework, dynamic XIP, which can resolve the performance degradation of using XIP. Using a constrained RAM cache, dynamic XIP can dynamically change XIP region according to page access pattern to reduce performance degradation in execution time or energy consumption resulting from native XIP problem. The proposed framework consists of a page profiler gathering applications' memory access pattern using PMU and an XIP manager deciding that a page is accessed whether in main memory or in flash memory. The proposed framework is implemented and evaluated in Linux kernel. Our evaluation shows that our framework can reduce execution time at most 25% and energy consumption at most 22% compared with using XIP-only case adopted in general mobile embedded systems. Moreover, the evaluation shows that in execution time and energy consumption, our modified LRU algorithm with code page filters can reduce more than at most 90% and 80% respectively compared with applying just existing LRU algorithm to dynamic XIP.

• Journal of Information Processing Systems
• /
• v.17 no.1
• /
• pp.191-202
• /
• 2021

In an Internet of Things (IoT)-configured system, each device executes on-chip software. Recent IoT devices require fast execution time of complex services, such as analyzing a large amount of data, while maintaining low-power computation. As service complexity increases, the service requires high-performance computing and more space for embedded space. However, the low performance of IoT edge devices and their small memory size can hinder the complex and diverse operations of IoT services. In this paper, we propose a remote on-demand software code execution unit using the cloudification of on-chip code memory to accelerate the program execution of an IoT edge device with a low-performance processor. We propose a simulation approach to distribute remote code executed on the server side and on the edge side according to the program's computational and communicational needs. Our on-demand remote code execution unit simulation platform, which includes an instruction set simulator based on 16-bit ARM Thumb instruction set architecture, successfully emulates the architectural behavior of on-chip flash memory, enabling embedded devices to accelerate and execute software using remote execution code in the IoT environment.