• Title/Summary/Keyword: 2D PE

Search Result 166, Processing Time 0.027 seconds

Power-Efficient DCNN Accelerator Mapping Convolutional Operation with 1-D PE Array (1-D PE 어레이로 컨볼루션 연산을 수행하는 저전력 DCNN 가속기)

  • Lee, Jeonghyeok;Han, Sangwook;Choi, Seungwon
    • Journal of Korea Society of Digital Industry and Information Management
    • /
    • v.18 no.2
    • /
    • pp.17-26
    • /
    • 2022
  • In this paper, we propose a novel method of performing convolutional operations on a 2-D Processing Element(PE) array. The conventional method [1] of mapping the convolutional operation using the 2-D PE array lacks flexibility and provides low utilization of PEs. However, by mapping a convolutional operation from a 2-D PE array to a 1-D PE array, the proposed method can increase the number and utilization of active PEs. Consequently, the throughput of the proposed Deep Convolutional Neural Network(DCNN) accelerator can be increased significantly. Furthermore, the power consumption for the transmission of weights between PEs can be saved. Based on the simulation results, the performance of the proposed method provides approximately 4.55%, 13.7%, and 2.27% throughput gains for each of the convolutional layers of AlexNet, VGG16, and ResNet50 using the DCNN accelerator with a (weights size) x (output data size) 2-D PE array compared to the conventional method. Additionally the proposed method provides approximately 63.21%, 52.46%, and 39.23% power savings.

Effect of Duration Treated on Reproductive Performance in Holstein Cows Treated with Progesterone with and without Estradiol (인공수정 후 외인성 Progesterone과 Estradiol 병용 및 Progesterone 단독 투여 기간이 젖소의 번식 성적에 미치는 영향)

  • Baek, K.S.;Lee, W.S.;Park, S.B.;Park, S.J.;Kim, H.S.;Lee, H.J.;Ki, K.S.;Kang, S.J.;Ahn, B.S.
    • Journal of Embryo Transfer
    • /
    • v.22 no.3
    • /
    • pp.195-198
    • /
    • 2007
  • This study was carried out to investigate the effects of hormones treated (progesterone, and combination of progesterone and estradiol) on the reproductive performance of dairy cattle. The intravaginal CIDR was inserted in the vagina of cow on day 14 (P-7, PE-7), 16 (P-5, PE-5) and 18 (P-3, PE-3) post-insemination to stimulate progesterone secretion with (PE) and without estradiol(P). The CIDR was removed on 7 (P-7, PE-7), 5 (P-5, PE-5) and 3 (P-3, PE-3) days after. The cows for control group were not treated with CIDR or hormones. Conception rate, the duration needed for conception after AI and the number of AI services needed for conception were measured. Conception rate in control, CIDR without estradiol (P-7, P-5 and P-3) and CIDR treated (PE-7, PE-5 and PE-3) cows were 15.8%, 40.0%, 31.5%, 28.6%, 62.5%, 50.0% and 0%, respectively. The days needed for conception after AI in control, P-7, P-5, p-3, PE-7, PE-5 and PE-3 were 64.9, 63.0, 59.1, 8.0, 0.0, 18.9 and 83.3 days, respectively. The days needed for conception in cows treated with CIDR (PE-7) has shown longer than control and PE-3 (p<0.05). The number of AI services needed for conception in control, CIDR (PE-7, PE-5 and PE-3) and CIDR without estradiol treated cows (P-7, P-5 and P-3) were 27, 2.3, 1.9, 1.3, 1.0, 1.7 and 2.8 times, respectively. The number of AI services needed for conception have shown significantly lower in PE-7 compared to control and PE-3.

Effects of Phytezyme Supplementation on the Growth Performance and Nutrient Digestibility in Growing Pigs

  • Hong, J.W.;Kim, I.H.;Kwon, O.S.;Lee, S.H.;Bae, H.D.;Kang, S.J.;Yang, U.M.
    • Asian-Australasian Journal of Animal Sciences
    • /
    • v.14 no.10
    • /
    • pp.1440-1443
    • /
    • 2001
  • Forty eight $Duroc{\times}Yorkshire{\times}Landrace$ pigs ($15.2{\pm}0.3kg$ average initial BW) were used conducted to evaluate the effects dietary phytase (Phytezyme, WOOJIN, Co. Ltd) supplementation on the growth performance and nutrient digestibility in growing pigs. There were three pigs per pen and four pens per treatment. Treatments were 1) Con (corn-wheat-SBM), 2) PE0.1 (low-P iet+0.1% Phytezyme), 3) PE0.2 (Iow-P diet+0.2% Phytezyme), 4) PE0.3 (low-P diet+0.3% Phytezyme). During d 0 to 28, average daily gain was not significantly different among the treatments. Pigs fed PE0.3 diet significantly decreased (p<0.05) their average daily feed intake compared to pigs fed Con diet. Also, gain/feed in pigs fed PE0.1 and PE0.3 diet was improved (p<0.05) compared to pigs fed Con and PE0.2 diet (p<0.05). For d 28 to 56, pigs fed PE0.2 diet grew significantly faster (p<0.05) than pigs fed Con and PE0.1 diet. Gain/feed was greater (p<0.05) for PE0.2 and PE0.3 treatments than for Con. For overall period, average daily gain was increased (P<0.05) by the addition of 0.2% Phytezyme compared with Con. Gain/feed was significantly improved (p<0.05) by supplementation of PE0.3. Pigs fed PE0.2 showed increased DM and N digestibilities compared to pigs fed other treatments. Supplemented diets PE0.2 and PE0.3 improved (p<0.05) the apparent digestibility of Ca and P compared to other treatments. In conclusion, the results obtained from these feeding trials suggest that the Phytezyme supplementation of diets for growing-finishing pigs had improved growth performance and nutrient availability.

A design and implementation of high-performance 2D PE architecture in H.264 Motion Estimation (H.264 움직임 추정의 고속 2D PE 아키텍쳐 설계 및 구현)

  • Lee, Kyung-Ho;Kong, Jin-Hyeung
    • Proceedings of the IEEK Conference
    • /
    • 2008.06a
    • /
    • pp.405-406
    • /
    • 2008
  • This paper presents a high performance 2D PE architecture for H.264 Motion Estimation(ME). While existing 2D PE architectures reuse the overlapped data of adjacent search windows scanned in 1 or 3-way, the new architecture scan adjacent windows and multiple paths instead of single raster and zigzag scanning of adjacent windows in 4 way(up,down,left,right). By reducing the redundant access factor by 1.4, the new 4-way search window improve the memory bandwidth by 70-58% compared with 1/3-way search window. With Altera Stratix-III implementation, the high performance 2D PE architecture deals with SD ($720{\times}480$) video of 2 reference frame, $48{\times}48$ search area and $16{\times}16$ macroblock by 30fps at 97.1MHz.

  • PDF

Macroalkoxyamines and macroRAFT agents based on polyethylene for the syntheses of polyolefin based polar block copolymers

  • Lopez R. Godoy;Boisson C.;D'Agosto F.;Spitz R.;Boisson F.;Gigmes D.;Bertin D.
    • Proceedings of the Polymer Society of Korea Conference
    • /
    • 2006.10a
    • /
    • pp.377-377
    • /
    • 2006
  • Alkoxyamine and thiocarbonyl thio end capped polyethylene (PE) chains were synthesized using a direct and simple approach consisting in reacting di(polyethylenyl)magnesium (PE-Mg-PE) chains with a range of nitroxides and disulfides of thiocarbonyl thio compounds. PE-Mg-PE compounds were prepared by a catalyzed chain growth reaction of ethylene on nbutyloctylmagnesium (BOMg) with a neodymocene complex $(C_{5}Me_{5})_{2}NdCl_{2}Li(OEt_{2})_{2}$. Complete characterizations confirm the introduction of the desired end groups. The controlled radical polymerization (NMP and RAFT) of butyl acrylate mediated by these functional polyethylenes was successful.

  • PDF

Influences of Plant Extract Supplementation on Performance and Blood Characteristics in Weaned Pigs

  • Hong, J.W.;Kim, I.H.;Kwon, O.S.;Min, B.J.;Lee, W.B.;Shon, K.S.
    • Asian-Australasian Journal of Animal Sciences
    • /
    • v.17 no.3
    • /
    • pp.374-378
    • /
    • 2004
  • One hundred and twenty crossbred pigs [(Duroc$\times$Yorkshire)$\times$Landrace] were used to determine the effects of plant extract (PE) supplementation on performance and blood characteristics in weaned pigs fed a corn-dried whey-SBM based diet. Treatments were 1) NC (antibiotic free basal diet), 2) PC (NC diet+100 ppm apramycin and 100 ppm oxytetracycline), 3) PE 0.1 (NC diet+0.1% plant extract), 4) PE 0.2 (NC diet+0.2% plant extract) and 5) PE+AB (PC diet+0.1% plant extract). Through the entire experimental period, ADG of pigs fed PC (300 vs. 281 g/d), PE 0.2 (310 vs. 281 g/d) and PE+AB (306 vs. 281 g/d) diets was higher than that of pigs fed NC diet (p<0.05). However, no differences were found among the treatments for ADFI and gain/feed. At day 2 after the onset of the experiment, fecal consistency score of pigs fed PC, PE 0.1, PE 0.2 and PE+AB diets was lower than that of pigs fed NC diet. There were no significant differences in red blood cell, white blood cell, lymphocytes, neutrophils and monocytes concentrations of blood among the treatments. In conclusion, PE can be used to replace antibiotics in diets for weaned pigs without negative affects on performance. Optimal PE levels seemed to be 0.2% and the results obtained point out to a synergic effect of the combination of PE and antibiotic on performance in weaned pigs.

4-way Search Window for Improving The Memory Bandwidth of High-performance 2D PE Architecture in H.264 Motion Estimation (H.264 움직임추정에서 고속 2D PE 아키텍처의 메모리대역폭 개선을 위한 4-방향 검색윈도우)

  • Ko, Byung-Soo;Kong, Jin-Hyeung
    • Journal of the Institute of Electronics Engineers of Korea SD
    • /
    • v.46 no.6
    • /
    • pp.6-15
    • /
    • 2009
  • In this paper, a new 4-way search window is designed for the high-performance 2D PE architecture in H.264 Motion Estimation(ME) to improve the memory bandwidth. While existing 2D PE architectures reuse the overlapped data of adjacent search windows scanned in 1 or 3-way, the new window utilizes the overlapped data of adjacent search windows as well as adjacent multiple scanning (window) paths to enhance the reusage of retrieved search window data. In order to scan adjacent windows and multiple paths instead of single raster and zigzag scanning of adjacent windows, bidirectional row and column window scanning results in the 4-way(up. down, left, right) search window. The proposed 4-way search window could improve the reuse of overlapped window data to reduce the redundancy access factor by 3.1, though the 1/3-way search window redundantly requires $7.7{\sim}11$ times of data retrieval. Thus, the new 4-way search window scheme enhances the memory bandwidth by $70{\sim}58%$ compared with 1/3-way search window. The 2D PE architecture in H.264 ME for 4-way search window consists of $16{\times}16$ pe array. computing the absolute difference between current and reference frames, and $5{\times}16$ reusage array, storing the overlapped data of adjacent search windows and multiple scanning paths. The reference data could be loaded upward and downward into the new 2D PE depending on scanning direction, and the reusage array is combined with the pe array rotating left as well as right to utilize the overlapped data of adjacent multiple scan paths. In experiments, the new implementation of 4-way search window on Magnachip 0.18um could deal with the HD($1280{\times}720$) video of 1 reference frame, $48{\times}48$ search area and $16{\times}16$ macroblock by 30fps at 149.25MHz.

Sequential Percentile Estimation for Sequential Steady-State Simulation (순차적 시뮬레이션을 위한 순차적인 Percentile 추정에 관한 연구)

  • Lee, Jong-Suk;Jeong, Hae-Duck
    • The KIPS Transactions:PartD
    • /
    • v.10D no.6
    • /
    • pp.1025-1032
    • /
    • 2003
  • Percentiles are convenient measures of the entire range of values of simulation outputs. However, unlike means and standard deviations, the observations have to be stored since calculation of percentiles requires several passes through the data. Thus, percentile (PE) requires a large amount of computer storage and computation time. The best possible computation time to sort n observations is (O($nlog_{2}n$)), and memory proportional to n is required to store sorted values in order to find a given order statistic. Several approaches for extimating percentiles in RS(regenerative simulation) and non-RS, which can avoid difficulties of PE, have been proposed in [11, 12, 21]. In this paper, we implemented these three approaches known as : leanear PE, batching PE, spectral $P^2$ PE in the context of sequential steady-state simulation. Numerical results of coverage analysis of these PE approachs are present.

New systolic arrays for computation of the 1-D and 2-D discrete wavelet transform (1차원 및 2차원 이산 웨이브렛 변환 계산을 위한 새로운 시스톨릭 어레이)

  • 반성범;박래홍
    • Journal of the Korean Institute of Telematics and Electronics S
    • /
    • v.34S no.10
    • /
    • pp.132-140
    • /
    • 1997
  • This paper proposes systolic array architectures for compuataion of the 1-D and 2-D discrete wavelet transform (DWT). The proposed systolic array for compuataion of the 1-D DWT consists of L processing element (PE) arrays, where the PE array denotes the systolic array for computation of the one level DWT. The proposed PE array computes only the product terms that are required for further computation and the outputs of low and high frequency filters are computed in alternate clock cycles. Therefore, the proposed architecuter can compute the low and high frequency outputs using a single architecture. The proposed systolic array for computation of the 2-D DWT consists of two systolic array architectures for comutation of the 1-D DWT and memory unit. The required time and hardware cost of the proposed systolic arrays are comparable to those of the conventional architectures. However, the conventional architectures need extra processing units whereas the proposed architectures fo not. The proposed architectures can be applied to subband decomposition by simply changing the filter coefficients.

  • PDF

Measurement of Liquid Entry Pressure of PE and PVDF Hollow Fiber Membranes in Membrane Distillation Process (막증류 공정에서 PE 및 PVDF 중공사막의 액체투과압력 측정에 관한 연구)

  • Min, Ji Hee;Park, Min Soo;Kim, Jinho
    • Membrane Journal
    • /
    • v.25 no.3
    • /
    • pp.216-222
    • /
    • 2015
  • The method measuring LEP (liquid entry pressure) was optimized to evaluate the wettabilities of hydrophobic membranes which might affect long-term durability of membrane in MD (membrane distillation) process. Conductivity of the permeate was monitored to measure the LEPs of PE (polyethylene) and PVDF (polyvinylidene di-fluoride) hollow fiber membranes from highly concentrated synthetic feed water of 20 wt% NaCl. Holding time over 5 min and the ratio of membrane area to the tank volume more than $10m^2/m^3$ were required to ensure LEP value.