0% Complete
صفحه اصلی
/
ششمین کنفرانس بین المللی میکروالکترونیک ایران
Comparison between Hardware/Software Co-design of RiscV Vector and Scalar Implementation of Deep Neural Networks
نویسندگان :
Seyed Kian Mousavikia
1
Morteza Mousazadeh
2
1- دانشگاه ارومیه
2- دانشگاه ارومیه
کلمات کلیدی :
Deep Neural Networks،Field Programmable Gate Array،Hardware/Software Co-Design،Parallel Processing،RiscV،Vector Co-processor
چکیده :
This paper compares a hardware/software co-design of a RiscV vector with a RiscV scalar implementation of a deep neural network (DNN). For the vector implementation, all building blocks of a DNN are vectorized and written in vector intrinsic coding format. Focusing more on the convolution function as the main source of the latency, this function is written in a special parallel processing-favor method in the vector intrinsic level to boost execution speed. For the comparison, a sample scalar RiscV core is selected and paired with a vector-based RiscV co-processor. Also, the same sample DNN is implemented only on the scalar processor to demonstrate the speedup better. The system was implemented and tested on a field-programmable gate array (FPGA). As a result, the vector implementation outperformed the scalar version by a factor of 3 in terms of latency by only negligibly increasing the utilized sources on the FPGA.
لیست مقالات
لیست مقالات بایگانی شده
Three-Phase Bidirectional Integrated Meter Using Hall Effect Sensor in 0.5um CMOS Process
Shayan Anzali - Morteza Mousazadeh - Khayrollah Hadidi
Design of a High Voltage Common Mode Resilient FlexRay Receiver in 180nm/5V CMOS Process
Hamid Sadat Mansoury - Saeed Saeedi - Mojtaba Atarodi
Base Transit Time Investigation of InP/InGaAs HBT Optoelectronic Mixer Using Different Base Doping Profiles
Hassan Kaatuzian - Mehrdad Ghasemi - Mahdi NoroozOliaei
ساختار گرافنی جاذب کامل چند باندی تراهرتزی با قابلیت تنطیم فرکانس و مستقل از قطبش
یوسف رفیق ایرانی - جواد جاویدان - حمید حیدرزاده
A Bootstrapped Switch Based Efficient CMOS Full-Wave Active Rectifier for Biomedical Implants
Mahmood Alibakhshi - Farshad Gozalpour - Yarallah Koolivand
A New Design for 1.75 to 2.55 GHz GaN Power Amplifier with More Than 40 dBm Output Power and 12 dB Maximum Gain
Marzieh Chegini - Mahmoud Kamarei - HojjatAllah Nemati
Design of a High-Efficiency Deep Bias Class-AB Power Amplifier With 70% PAE at P1dB
Fazel Ziraksaz - Alireza Hassanzadeh
Characteristics of Electrophoretic Graphene-oxide (GO) Film on Permalloy (Ni-Fe) Thin Film Sputtered on Copper Substrate
Ali Rezaei
Design of All-Passive Harmonic-Rejection N-Path Filter
Vahid Moradino - Esmaeil Najafiaghdam
Quantum Machine Learning Acceleration with Quantum Control Techniques
Sara Mahmoudi Rashid
بیشتر
ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 41.5.5