» IPM HPC Laboratory

Jan 31

Winner in Memocode 2015 Design Contest

13th ACM-IEEE International Conference on Formal Methods and Models for System Design
The University of Texas at Austin
September 21-23, 2015

The winners of 2015 MEMOCODE Design Contest:

Rank in Best Cost-Normalized Performance and Highest Performance class: 2- IPM-HPC, 1- Tokyo University

IPM-HPC Team members: Armin Ahmadzadeh, Ehsan Montahaie, Milad Ghafouri, Reza Mirzaei, Saied Rahmani, Farzad Sharif Bakhtiar, Mohsen Gavahi, Rashid Zamanshoar, Hanie Ghasemi, Kianoush Jafari, Saeid Gorgin

Institute for Research in Fundamental Sciences (IPM), Iran

Contest Problem: As in previous years, MEMOCODE’15 will include a design contest, which will pose a computational challenge that participants may solve using hardware or software on FPGAs, GPUs, and CPUs. The conference will sponsor at least one prize with a monetary award for the contest winners. the 2015 challenge is continuous skyline computation that given dataset is not constant and it changes over time. the aim of this contest is to implement a system to efficiently compute the continuous skyline of a large dynamic dataset.

IPM-HPC Solution:  Our method makes exhaustive use of CPU and minimizes memory access. we present an efficient parallel continuous Skyline approach. In our suggested method, the dataset points are sorted and pruned based on Manhattan distance. Moreover, we use several optimization methods to optimize memory usage in comparison with naïve implementation. In addition, besides the applied conventional parallelization methods, we partition the time steps based on the number of available cores. The experimental results for a data set that contains 800k points with 7 dimensions show considerable speedup.for more information see the paper.

Jan 13

Winner in MEMOCODE 2014 Design Contest

Winner in Memocode 2014-Hardware Contest

Switzerland, EPFL in Lausanne
Rank in Best Cost-Normalized Performance class: 1 – IPM-HPC-MultiCore (winner), 2 and 3- IPM-HPC-ManyCore (GPU), 4- Iowa State University (USA)
Rank in Highest Performance class: 2- IPMH-HPC, 1- Iowa State University (USA) (winner)
Our team defended its previous years rank in performance/cost class.

IPM-HPC Team members:  Armin Ahmadzadeh, Reza Mirzaei, Hatef Madani, Mohammad Shobeiri, Mahsa Sadeghi, Mohsen Gavahi, Kianoush Jafari, Mohsen Mahmoudi Aznaveh, and Saeid Gorgin.

Contest Problem:The 2014 problem for The MEMOCODE hardware contest problem is k-Nearest Neighbor search using the Mahalanobis distance metric. Given a data set of points in multidimensional space, the goal is to find the k points that are nearest to any given point in that space.

IPM-HPC Solution: Regarding to the MEMOCODE chair email, there were 10 excellent solutions submitted, utilizing a range of algorithmic as well as system-level optimizations, targeting FPGA, GPU as well as multi-core platforms. Our method makes exhaustive use of CPU and minimizes memory access. This method is the winner of Best Cost-Normalized Performance of MEMOCODE contest design 2014 and is 616X faster than other implementation of the contest. See this paper for more detail.

 

Winner in Memocode 2014-Software Contest

Switzerland, EPFL in Lausanne
Rank in Software contest: 2 – IPM-HPC, 1-U.S. Army Research, Development and Engineering (USA) – Winner

IPM-HPC Team members:Nariman Eskandari, Hatef Madani, Armin Ahmadzadeh, Saied Gorgin, Mohsen Mahmoudi Aznaveh.

Contest Problem:The 2014 problem for The MEMOCODE software contest problem is to make the emulator run even faster on the Raspberry PI by proposing software solutions. These contest used specific emulator for Space Invaders game written for 8080 processor.

IPM-HPC Solution: Regarding to the MEMOCODE software contest report, there were three contestants were successful in completing this software design contest. These contestants employed a variety of techniques to both discover and optimize performance bottlenecks in the emulator. Our method is focused in optimizing function calls that were frequently invoked, and re-organizing data-structures to better match the underlying Raspberry Pi’s hardware architecture. These approach is 2.5 times faster over the conventional implementation of the contest. See this paper for more detail.

May 19

کارگاه دو روزه برنامه‌نویسی پردازنده‌های چند هسته‌ای

 

عناوین
مقدمه ای بر معماری کامپیوتر و معماری موازی
زبان های برنامه نویسی پردازنده های چند هسته ای
تکنیک های پردازش موازی در داخل هسته
مدل برنامه نوسی پردازنده های گرافیکی
کارگاه برنامه نویسی CUDA، OpenMP و OpenCl
 مباحث پیشرفته در برنامه نویسی پردازنده های چند هسته ای
تاریخ و محل برگزاری
تاریخ:  چهارشنبه و پنجشنبه – 15 و 16 مرداد (9:00 – 17:00)
تهران، خيابان شهيد لواساني (ساختمان فرمانيه)، بعد از برج کوه نور

توجه: ساعت شروع کارگاه در روز چهارشنبه ساعت 8:30 دقیقه می باشد.

پیش ثبت نام
در صورت تمایل به شرکت در کارسوق باید در آدرس زیر مشخصات خود را جهت پیش ثبت نام ارسال کنید:

http://hpc.ipm.ac.ir/

به علت ظرفیت محدود کارسوق، کمیته برگزاری متناسب با امکانات موجود، تعدادی از متقاضیان را برای ثبت نام نهایی انتخاب خواهد کرد. این انتخاب از طریق ایمیل به اطلاع متقاضیان خواهد رسید و ایشان باید برای تکمیل ثبت نام هزینه تعیین شده را پرداخت کنند.
هزینه ثبت نام دانشجویی 1.000.000 ریال (معادل یکصد هزار تومان) و هزینه ثبت نام آزاد 2.500.000 ریال است.

اطلاعات حساب (جهت واریز وجه ثبت نام) 

شماره حساب : 2172149001004
بانک عامل : ملی
شعبه : نیاوران
به نام: پژوهشکده دانش های بنیادی (IPM)
عنوان پرداخت : کارسوق برنامه نویسی چند هسته ای
 مهلت پرداخت : دوشنبه 93/05/13
توجه : لطفا نسبت به پرداخت هزینه با توجه به وضعیت خود اقدام فرمایید و تصویر رسید آن را به آدرس hpc@ipm.ir تا آخر وقت اداری روز دوشنبه مورخ  93/5/13 ارسال نمایید.

همچنین اصل رسید و در صورت دانشجو بودن، کارت دانشجویی خود را در روز برگزاری کارسوق به همراه داشته باشید.

مشخصات تماس
تهران، خيابان شهيد لواساني (ساختمان فرمانيه)، بعد از برج کوه نور
وب سایت:http://hpc.ipm.ac.ir
رایانامه:hpc (- at -) ipm .ir
شماره تماس: 982124509409+

Sep 19

Memocode Design Contest 2013 Winner

IPM-HPC laboratory won the Memocode Design Contest 2013 for the third time. Our team not only defended its previous years rank in performance/cost class, but also we won the performance class. The 2013 problem for the MEMOCODE Design Contest was part of the stereo matching problem: Given a stereo image pair, the challenge was to infer the depth information for each pixel in the image.

Regarding to the email that we have received from the design Contest chair, there were 8 excellent solutions submitted to the contest this year, utilizing a range of algorithmic as well as system-level optimizations, targeting FPGA, GPU as well as multi-core platforms. The top performing solutions, including ours, were provided the judging image to help the judges in making the final decision.

Based on the run-time, cost, and inference accuracy, IPM Team’s solution has been selected as the all-around winner.

IPM-HPC Team members: Armin Ahmadzadeh, Hatef Madani, Kianoush Jafari,  Shervin Daneshpajouh, Saeid Gorgin

کارگاه دو روزه برنامه‌نویسی مقدماتی پردازنده‌های چند هسته‌ای

عناوین

مقدمه ای بر معماری کامپیوتر و معماری موازی

زبان های برنامه نویسی پردازنده های چند هسته ای

تکنیک های پردازش موازی در داخل هسته

مدل برنامه نوسی پردازنده های گرافیکی

استفاده از پردازنده های گرافیکی در MATLAB

 کارگاه برنامه نویسی CUDA  و OpenMP

تاریخ و محل برگزاری

تاریخ: پنجشنبه و جمعه – 19 و 20 بهمن آدرس: تهران، نیاوران، میدان شهید باهنر، پژوهشگاه دانش های بنیادی

مشخصات تماس

وب سایت: http://cs.ipm.ac.ir/tdwmc2013

رایانامه:hpc (- at -) ipm .ir  شماره تماس: +98 21 22814933


 

 

 

May 16

Memocode Design Contest 2012 Winner

IPM-HPC laboratory, defended its previous rank and won the Memocode design contest for the second time. The 2012 problem for the MEMOCODE Design Contest was part of the DNA sequence alignment problem: exact substring matching. The challenge was to efficiently locate millions of 100-base-pair short read sequences in a 3-million-base-pair reference genome.

IPM-HPC team provided two solutions: The first one is based on BWT data structure and the second one is based on Hash-indexing.  Both of these methods use CPU-multicore and PThread technology. The Hash indexing method was one of the two winners of the contest. The method is able to align 1.2 GB of short read sequences in only 4 seconds. The hash-based solution become the winner of the normalized class of the contest. The hpc-team members of this year were: Aryan Arbabi,Milad Gholami, Mojtaba Varmazyar, Shervin Daneshpajouh. See details in the MEMOCODE page.

Mar 05

Cloud Computing Workshop

HPC Researchers Group presented their work in a two day workshop, on 1,2 March 2012

Thursday (1 March) Talks:

  1. Introduction to Cloud Computing
  2. An Overview of Virtualization
  3. Service Models/Layers of Cloud Computing
  4. Amazon EC2 (Demo)
  5. Data Management in Cloud Systems

Friday (2 March) Talks:

  1. HPC in Cloud Computing
  2. Pricing and Scalability
  3. Data Management in Cloud (Demo)
  4. Cloud Computing: Cost Models
  5. Challenges and Opportunities

Join the Forum discussion on this post

Join the Forum discussion on this post

Oct 25

GPU Talks by Shiraz University Researchers

Shiraz University GPU Research Group presented their work in a two day workshop, on 21,22 October 2011.

Thursday (21 October) Talks:
1) MapReduce Programming Model
2) Surena: A Collaborative CPU-GPU Framework in Hadoop-based MapReduce Clusters
3) Soren: An adaptive MapReduce framework on GPUs
4) Performance Analysis on GPUs
Friday (22 October) Talks:
1) Genetic Algorithms on GPUs
2) Hands-on experience in Hadoop-based MapReduce applications
3) Hands-on experience in GPU Performance Analysis

Oct 13

Talks on Linux Kernel, Kickstart and Hadoop

Some Talks were held in HPC Lab on Thursday 6 October 2011.

Talks Titles:

1) Cluster Zero (Automated Installation by Kickstart)
2) Linux kernel
3) Tosan Web Interface for Cluster Services
4) Hadoop

Aug 21

CLOUDS Research Group

You Could see “The Cloud Computing and Distributed Systems (CLOUDS) Research Group” Home page here.

Older posts «

Powered by PHP Webmasterpoint.org