• Title/Summary/Keyword: Research

Search Result 342,087, Processing Time 0.193 seconds

AB9: A neural processor for inference acceleration

  • Cho, Yong Cheol Peter;Chung, Jaehoon;Yang, Jeongmin;Lyuh, Chun-Gi;Kim, HyunMi;Kim, Chan;Ham, Je-seok;Choi, Minseok;Shin, Kyoungseon;Han, Jinho;Kwon, Youngsu
    • ETRI Journal
    • /
    • v.42 no.4
    • /
    • pp.491-504
    • /
    • 2020

  • We present AB9, a neural processor for inference acceleration. AB9 consists of a systolic tensor core (STC) neural network accelerator designed to accelerate artificial intelligence applications by exploiting the data reuse and parallelism characteristics inherent in neural networks while providing fast access to large on-chip memory. Complementing the hardware is an intuitive and user-friendly development environment that includes a simulator and an implementation flow that provides a high degree of programmability with a short development time. Along with a 40-TFLOP STC that includes 32k arithmetic units and over 36 MB of on-chip SRAM, our baseline implementation of AB9 consists of a 1-GHz quad-core setup with other various industry-standard peripheral intellectual properties. The acceleration performance and power efficiency were evaluated using YOLOv2, and the results show that AB9 has superior performance and power efficiency to that of a general-purpose graphics processing unit implementation. AB9 has been taped out in the TSMC 28-nm process with a chip size of 17 × 23 ㎟. Delivery is expected later this year.

  • https://doi.org/10.4218/etrij.2020-0134 인용 PDF KSCI

Recent Progress in Organic Thin Film Transistor on the Plastic Substrates

  • Suh, Kyung-Soo;Kang, Seung-Youl;Ahn, Seong-Deok;Oh, Ji-Young;You, In-Kyu;Kim, Gi-Hyun;Baek, Kyu-Ha;Kim, Chul-Am;Hwang, Chi-Sun;KoPark, Sang-Hee;Yang, Yong-Suk;Chung, Sung-Mook;Lee, Jeong-Ik;Do, Lee-Mi;Chu, Hye-Yong;Kang, Kwang-Yong
    • 한국정보디스플레이학회:학술대회논문집
    • /
    • 2005.07a
    • /
    • pp.61-63
    • /
    • 2005

  • Pentacene based OTFT on PC and PES plastic substrates have been fabricated in a scale of 5 inches. We could get a small OTFT device enough to be applicable for AMOLED by acquiring the at least misalignment margin through a contact aligner. And also we could find out the degradation of device parameter through the integration processes and improve the properties by using a buffer layer as an etch stopper in an active patterning. Through these, the mobility of device is more than about $0.2cm^2/Vs$ and $I_{on}/I_{off}$ is higher than $10^5$.

  • PDF