• Journal of Advanced Navigation Technology
• /
• v.13 no.4
• /
• pp.509-514
• /
• 2009

In this paper, we designed and analyzed 256GB DRAM-based SSD storage using DDR1 memory and PCI-e interface. SSD is a storage system that uses DRAM or NAND Flash as primary storage media. Since the SSD read and write data directly to memory chips, which results in storage speeds far greater than conventional magnetic storage devices, HDD. Architecture of the proposed SSD system has performance of high speed data processing duo to use multiple RAM disks as primary storage and PCI-e interface bus as communication path of RAM disks. We constructed experimental system with UNIX, Windows/Linux server, SAN Switch, and Ethernet Switch and measured IOPS and bandwidth of proposed SSD using IOmeter. In experimental results, it has been shown that IOPS, 470,000 and bandwidth,800MB/sec of the DDR-1 SSD is better than those of the HDD and Flash-based SSD.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.10a
• /
• pp.386-389
• /
• 2013

In this paper, for signal integrity analysing high-speed signal between a processor and a DDR2 memory, transient analysis is done and eye diagrams are generated using IBIS models of IC chips and S-parameters of transmission line. From the eye diagrams of such high-speed signals as DQ, DQS/DQSb, Clock, Address and Control, signal quality is assured through measuring timing and voltage margins during setup and hold times and verifying that the over-/under-shoot and the cross points of differential signals satisfy their specifications.

• Journal of Broadcast Engineering
• /
• v.23 no.5
• /
• pp.606-613
• /
• 2018

This study is concerned with implementing an external memory expansion device for large-scale image processing. It consists of an external memory adapter card with a PCI(Peripheral Component Interconnect) Express Gen3 x8 interface mounted on a graphics workstation for image processing and an external memory board with external DDR(Dual Data Rate) memory. The connection between the memory adapter card and the external memory board is made through the optical interface. In order to access the external memory, both Programmable I/O and DMA(Direct Memory Access) methods can be used to efficiently transmit and receive image data. We implemented the result of this study using the boards equipped with Altera Stratix V FPGA(Field Programmable Gate Array) and 40G optical transceiver and the test result shows 1.6GB/s bandwidth performance.. It can handle one channel of 4K UHD(Ultra High Density) image. We will continue our study in the future for showing bandwidth of 3GB/s or more.

• Journal of IKEEE
• /
• v.20 no.3
• /
• pp.241-250
• /
• 2016

AVM(Around View Monitoring) is a function of ADAS(Advanced Driver Assistance Systems), which provides a bird's eye view of the surroundings of a vehicle to the user. AVM systems require large bandwidth since they are composed of four input images and require real-time processing for vehicle-embedded environments. Also, the memory bandwidth requirement increases greatly when the resolution of the input data is higher. In this paper, we propose four basic hardware models of AVM systems. The models are decided by whether or not there is a valid data extraction module and an image processing purpose LUT generation module. We analyze the required bandwidth and hardware resource for each model. For verification of the proposed models, we implemented an AVM system using XC7Z045 FPGA and DDR3 memory for VGA and FHD resolution. All four of the proposed hardware model is executed below 33ms, which shows that it can operate in real-time.

• Journal of IKEEE
• /
• v.18 no.4
• /
• pp.509-516
• /
• 2014

CRC feature in a high-speed semiconductor memory devices such as DDR4/GDDR5 increases the data reliability. Conventional CRC method have a massive area overhead and long delay time. It leads to insufficient internal timing margins for CRC calculation. This paper, presents a CRC code method that provides error detection and a real-time matrix type CRC. If there are errors in the data, proposed method can alert to the system in a real-time manner. Compare to the conventional method(XOR 6 stage ATM-8 HEC code), the proposing method can improve the error detection circuits up to 60% and XOR stage delay by 33%. Also the real-time error detection scheme can improve the error detection speed to agerage 50% for the entire data bits(UI0~UI9).

• Journal of Korea Multimedia Society
• /
• v.14 no.3
• /
• pp.368-377
• /
• 2011

This paper explains the hardware structure of SURF(Speeded Up Robust Feature) based feature point extractor and its FPGA verification result. SURF algorithm produces novel scale- and rotation-invariant feature point and descriptor which can be used for object recognition, creation of panorama image, 3D Image restoration. But the feature point extraction processing takes approximately 7,200msec for VGA-resolution in embedded environment using ARM11(667Mhz) processor and 128Mbytes DDR memory, hence its real-time operation is not guaranteed. We analyzed integral image memory access pattern which is a key component of SURF algorithm to reduce memory access and memory usage to operate in c real-time. We assure feature extraction that using a Vertex-5 FPGA gives 60frame/sec of VGA image at 100Mhz.

• Journal of KIISE
• /
• v.44 no.9
• /
• pp.878-886
• /
• 2017

This paper discusses performance bottlenecks that may occur when executing high-performance computing MPI applications in the Intel's next generation many-core processor called Knights Landing(KNL), as well as effective resource allocation techniques to solve this problem. KNL is composed of a host processor to enable self-booting in addition to an existing accelerator consisting of a many-core processor, and it was released with a new type of on-package memory with improved bandwidth on top of existing DDR4 based memory. We empirically verified an improvement of the execution performance of multiple MPI applications and the overall system utilization ratio by studying a resource allocation method optimized for such new many-core processor architectures.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.11
• /
• pp.2285-2291
• /
• 2011

In this paper, H.264 high profile intra frame encoder, which integrates intra prediction, context-based adaptive variable length coding(CAVLC), and DDR2 memory control module, is proposed. The designed encoder can be operated in 440 cycle for one-macroblock. In order to verify the encoder function, we developed the reference C from JM 13.2 and verified the developed hardware using test vector generated by reference C. The designed encoder is verified in the FPGA (field programmable gate array) with operating frequency of 200 MHz for DMA (direct memory access), operating frequency of 50 MHz of Encoder module, and 25 MHz for VIM(video input module). The number of LUT is 43099, which is about 20 % of Virtex 5 XC5VLX330.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.37 no.7
• /
• pp.43-49
• /
• 2000

We proposed ADD(Alternate Directional Delay) circuit technique as the DLL(Delay Locked Loop) circuits which technique is established the data valid window(tDV) in DDR(Double Data Rate) Synchronous DRAM. This technique could be decrease area-overhead which it could generated bidirectional clock simultaneously using only one delay chain block. In this paper for high speed memory with relatively small size. This technique decreased area-overhead more 2 times than SMD(Synchronous Mirror Delay) technique. ADD technique has 50ps-140ps jitter and the operation frequency has 166MHz-66MHz range.(at 2.5V, TYP. condition)

• Journal of IKEEE
• /
• v.19 no.3
• /
• pp.288-294
• /
• 2015

CRC function has been built into the high-speed semiconductor memory device in order to increase the reliability of data for high-speed operation. Also, DBI function is adopted to improve of data transmission speed. Conventional CRC(ATM-8 HEC code) method has a significant amounts of area-overhead(~XOR 700 gates), and processing time(6 stage XOR) is large. Therefore it leads to a considerable burden on the timing margin at the time of reading and writing of the low power memory devices for CRC calculations. In this paper, we propose a CRC method for low cost and high speed memory, which was improved 92% for area-overhead. For low-cost implementation of the CRC scheme by the DBI function it was supplemented by data bit error detection rate. And analyzing the error detection rate were compared with conventional CRC method.