ASC16 Awards

  1. Champion: Huazhong University of Science and Technology
  2. Silver Winner: Shanghai Jiao Tong University
  3. e Prize: Huazhong University of Science and Technology
  4. Highest LINPACK: Zhejiang University
  5. Application Innovation Award: Nanyang Technological University, Singapore
  6. Application Innovation Award: Beihang University
  7. Application Innovation Award: National University of Defense Technology
  8. Application Innovation Award: Sun Yat-sen University
  9. Most Popular Team: Hong Kong Baptist University (WeChat)  Nanyang Technological University, Singapore (Twitter)

 

ASC16 Final Competition Notification

Dear ASC16 finalist:

Congratulations on getting into the final round of the ASC Student Supercomputer Challenge 2016! The competition will be held at Huazhong University of Science and Technology in Wuhan, China from April 18 to April 22, 2016. Please find below the agenda and more detailed information in the attachment. Feel free to ask techsupport@asc-events.org if you have any questions.

Date Time Content (1) Content (2)
April 17   Check-in
April 18 08:00-20:00 Announcement of contest rules, cluster building and testing The power consumption of the entire cluster should be less than 3000W DNN application debugging DNN shall be optimized on Tianhe-2
April 19 08:00-20:00 Cluster building and testing DNN application debugging
April 20 08:00-18:00 Performance testing of HPL, HPCG, MASNUM_WAM application Performance testing of DNN
April 21 08:00-18:00 Performance testing of ABINIT, Mystery  Application Performance testing of DNN(08:00-12:00)
April 22 08:30-12:15 Team Presentation
14:30-18:30 The Awards Ceremony

Equipment: The hardware platform will be provided by the Inspur Group (detailed configuration described in Appendix I). Every team is required to fill and submit Hardware Platform Equipment of the Final Stage (please refer to the attached "ASC16 Student Supercomputer Challenge --- Notifications about the Final Stage") to techsupport@asc-events.org of this contest by March 31.

Competition
: HPL, HPCG, MASNUM, Abinit, DNN, Mystery Application, Team Presentation (details, rules, and agenda in the attached guidebook)

Training: To help you better understand the rules and the platform of the final contest, we will hold a training program at Huazhong University of Science and Technology on March 18. The detailed agenda is in Appendix V. Trip expenses will not be covered. Please fill out the form below and send it to info@asc-events.org by March 13 if you plan to attend.

University Name Title Email Mobile
         
         
         
         
         
         

Sincerely,
ASC16 Committee

ASC16 Student Supercomputer Challenge
Final Competition Preview

Rules of the final stage

  1. Optimization methods that are only applicable to specific parameters or input data are strictly prohibited.
  2. If there are any modifications on the algorithm, the new algorithm must be mathematically equivalent to the original one.
  3. If any rules given above are violated, a score of zero will be given for the corresponding task.
    Note: when in doubt, a team needs to submit a query to the contest committee before the competition on whether a specific optimization method violates the rules, and a decision will be made by the evaluation committee before the competition. Otherwise, the team will have no chance to provide further explanations when its optimization method is ruled out by the evaluation committee during the competition.

Arrangement of the final stage

  1. Every team is required to fill and submit Hardware Platform Equipment of the Final Stage (Appendix I) to the technical supporting email (TechSupport@asc-events.org) of this contest by March 31; the committee will prepare the corresponding equipment for the teams to use in the final competition.
  2. During the preparation period of the finals, the committee will provide a remote testing platform for every team, please see Appendix III for details about the login information. 
  3. The ASC16 committee will provide the accommodation for every team (five students and one advisor) from April 17th to 23th, 2016, and the committee will provide a guidance session before the actual competition.
  4. Please contact the committee of the contest if there is any doubts about the rules: 

Appendix I

ASC16 Student Supercomputer Challenge
——Hardware Platform and Equipment of the Final Stage

Restriction of power consumption and hardware platform
Item Name Configuration
Server Inspur
NF5280M4

CPU: Intel Xeon E5-2650v3 x 2,2.3Ghz,10 cores
Memory: 16G x8,DDR4,2133Mhz
Hard disk: 1TB SATA x 1
Estimated power consumption:
E5-2650v3 TDP 105W,memory nominal 7.5W, hard disk
nominal 10W

Accelerator card XEON PHI-31S1P

Intel XEON PHI-31S1P (57 cores, 1.1GHz, 1003GFlops,
8GB GDDR5 Memory)
Estimated power consumption: 270W

HCA card FDR

Infiniband Mellanox Connect X®-3 HCA card, single port QSFP,
FDR IB MCX353A-FCBT
Estimated power consumption:9W

Switch GbE switch

10/100/1000Mb/s,24 ports Ethernet switch
Estimated power consumption:30W

FDR-IB switch

Switch X™ FDR Infiniband switch, 36 QSFP port
Estimated power consumption:130W

Cable Gigabit CAT6 cables

CAT6 copper cable, blue, 3m

Infiniband cable

Infiniband FDR optical fiber cable, QSFP port, working with
the Infiniband switch

Tips of power consumption: the power consumption of a single server above (excluding accelerators) is about 350W when running with full workload.


Appendix II

ASC16 Student Supercomputer Challenge
--Technical Regulation and Evaluation Criteria of the Final Stage

ARestrictions

  1. HPL、HPCG、MASNUM_WAM、ABINIT、Mystery Application:
    The power consumption must be under 3000W. Otherwise, no result will be accepted.
  2. DNN equipment specification:
    DNN shall be optimized on Tianhe-2 (see B and VI).

BPerformance Optimization (90 points)
I. HPL performance optimization (10 points):

  1. Platform requirement: The power consumption must be under 3000W. Otherwise, the current task result becomes invalid. 
  2. Goal: The highest performance is the goal while passing the correctness checking.
  3. Software downloading: http://www.netlib.org/benchmark/hpl/

II. Performance optimization of HPCG (10 points):

  1. Platform requirement: The power consumption must be under 3000W. Otherwise, the current task result becomes invalid.
  2. About run time: HPCG (version 3.0) runs must be at least 1800 seconds (30 minutes) as reported in the output file. The Quick Path option is not allowed.
  3. Software downloading: http://www.hpcg-benchmark.org/software/index.html

III. Performance optimization of MASNUM_WAM (15 points):

  1. Platform requirement: The power restriction of the test platform is 3000W. If the power consumption of system exceeds 3000W during the contest, the current task result becomes invalid. 
  2. Goal: The committee will announce multiple MASNUM_WAM workloads during the finals. Every team can only modify the options related to the parallel setting. Other modifications of the workloads are prohibited. The team needs to pass the correctness checking of each workload, and the goal is to achieve the shortest runtime of each workload.

IV. Performance optimization of ABINIT (15 points):

  1. Platform requirement: The power restriction of the test platform is 3000W. If the power consumption of system exceeds 3000W during the contest, the current task result becomes invalid. 
  2. Goal: The committee will announce several ABINIT workloads during the finals. Every team can only modify the options related to the parallel setting. Other modifications of the workloads are prohibited. The team needs to pass the correctness checking of each workload, and the goal is to achieve the shortest runtime of each workload
  3. ABINIT source code downloading: http://www.abinit.org/downloads/abinit-packages

V. Performance optimization of the Mystery Application (15 points):

  1. Platform requirement: The power restriction of the test platform is 3000W. If the power consumption of system exceeds 3000W during the contest, the current task result becomes invalid.  
  2. Goal: The committee will announce the Mystery Application software and the corresponding workloads on site to all the teams at the same time. Each team can then perform application compilation and optimization; every team can only modify the options related to the parallel setting. Other modifications of the workloads are prohibited. Every workload result is required to pass the correctness checking and to achieve the shortest runtime of all workloads.

VI. Performance optimization of the DNN program (25 points):

  1. Introduction of the contest: All teams are required to conduct performance optimization of the DNN program on MIC cluster. The committee will provide one or more workloads for the final competition. The goal is to achieve the highest performance of all the workloads.
  2. Platform: MIC cluster on Tianhe-2
    Configure: The cluster has eight MIC nodes. Each contains 2 CPUs (Intel(R) Xeon E5-2692 v2, 12 cores, 2.20GHz) and 3 MIC cards (Intel(R) Xeon Phi 31S1P, 57 cores, 1.1GHz, 8GB memory). The Tianhe-2 network is self-customized high speed interconnect system, bandwidth between nodes is 160Gbps. A specific realization of MPI for Tianhe-2 is provided. The compiling command is mpic++.
  3. Debug case download linkasc16.inspur.com:/home/public
    Correctness verification requirement of debug case: Accuracy ≥45.7
  4. Rules of the optimization:

VII. Evaluation method of the Applications:


Applications
Points Evaluation method
Performance Optimization
(90 points)
HPL 10

Let  be the actual performance of each team in which  is the maximum, the score P1 will be given as:

HPCG 10

P2 is calculated in the same way as P2 in HPL.

MASNUM_WAM 15

, where N is the number of workloads,  is the full score of the ith workload, the score P3 will be given as:

Where  is the runtime of the ith workload, and  is the minimum among all the participating teams.

ABINIT 15

P4 is calculated in the same way as P3 in MASNUM_WAM.

Mystery  Application 15

P5 is calculated in the same way as P3 in MASNUM_WAM.

DNN 25

Performance Total Points

CTeam Presentation (10 points)


Appendix III

ASC16 Student Supercomputer Challenge
-- Remote Test Platform for Preparing the Finals

I. Configuration of the remote testing platform
The platform consists of login/compiling nodes and cluster nodes. The detailed configuration is as follows:

Item Name Configuration Hosts

Server

Inspur NF5280M4 x 4

CPU:Intel Xeon E5-2680v3 x 2, 2.5Ghz, 12 cores

hostname:
cm1,
cm2,
cm3,
cm4

Memory: 16G x8, DDR4, 2133Mhz

Hard disk: 1T SATA  x 1

Item Name Configuration Host

Server

Inspur NF5280M4 x 1

CPU:Intel Xeon E5-2680v3 x 2, 2.5Ghz, 12 cores

hostname:
mc1

Memory: 16G x8, DDR4, 2133Mhz

Hard disk: 1T SATA  x 1

Accelerator card: Intel XEON PHI-5110P (60 cores, 1.053GHz, 1010GFlops, 8GB GDDR5 Memory)

Item Name Configuration Hosts

Server

Inspur NF5280M4 x 4

CPU:Intel Xeon E5-2680v3 x 2, 2.5Ghz, 12 cores

hostname:
cpu1,
cpu2,
cpu3,
cpu4

Memory: 16G x8, DDR4, 2133Mhz

 

Hard disk: 1T SATA  x 1

 

Network

 

Infiniband+Ethernet

 

Item Name Configuration Hosts

Server

Inspur NF5280M4 x 4

CPU:Intel Xeon E5-2680v3 x 2, 2.5Ghz, 12 cores

hostname: mic1,
mic2,
mic3,
mic4

Memory: 16G x8, DDR4, 2133Mhz

Hard disk: 1T SATA  x 1

Accelerator card: Intel XEON PHI-31S1P (57 cores,
1.1GHz, 1003GFlops, 8GB GDDR5 Memory)

Network

 

Infiniband+Ethernet

 


II. Instructions of remote testing platform

1. Login account:
Platform login IP address: asc16.inspur.com
The login account and password for users to log in ASC16 platform shall be configured by system administrator and will be notified to users through E-mail. Users shall change the password after first login. The login account will become automatically invalid after the contest is finished. The same account allows at most 2 users to log in at the same time.

2. Storage space: 
Each user will have 20GB storage space by default. Users shall make their own data backup and timely clear the disk space. In case the user needs to use a larger disk space, please contact the ASC administrator directly through techsupport@asc-events.org.

3. Principles for using the testing platform
As mentioned above, the login/compiling nodes shall be used for login, compiling, and doing single-node tests. Multi-node test shall be performed on the cluster nodes by submitting jobs.
For more detailed description, please refer to “ASC16 Instruction Manual of Remote Login Platform.doc” which sent to you by E-mail in Preliminary Contest.
Any problem encountered by users in using the ACS16 remote testing platform shall be sent to techsupport@asc-events.org, and the system administrator will provide the answer. 


Appendix IV

ASC16 Student Supercomputer Challenge
-- Tools

The ASC Committee has gathered two tools that you could use in the final competition:

OpenACC
OpenACC is an auto-parallelizing programming model to accelerate your application without significant programming effort. It uses high-level compiler directives to expose parallelism in the code and parallelizing compilers to build the code for a variety of parallel accelerators, for example GPU and multi-core CPU. OpenACC possesses the features of simple, powerful and portable and supports C/C++ and Fortran. You may download it at https://developer.nvidia.com/openacc. You will need to copy the attached 60-day license to the installation home directory.

Teye
The remote log-in platform provides the monitoring and scheduling software ClusterEngine, and the application features analyzing system Teye that can provide real-time monitoring for the resource utilization of the underlying hardware.
You can access http://asc16.inspur.com to see job status (job monitor and Teye information), and you can also see login time and power utilization information. Besides, you can download ClusterEngine and Teye installation package and user manual from asc16.inspur.com/home/public to make use of them in your local environment, in this case, you need to email techsupport@asc-events.org to apply for a trial license.

 
Appendix V

ASC16 Student Supercomputer Challenge
-- Training for Final Competition

Time      March 18, 2016
Venue    International Academic Exchange Center No. 8, Academic Lecture Hall on the 3rd Floor, Huazhong University of Science and Technology

Time Agenda
14:30-14:40 Speech by ASC16 Co-Host - HUST
14:40-14:50

Speech by Asia Supercomputer Community

14:50-15:20 ASC16 Final Competition Ground Preparation
15:20-15:30 ASC16 Top16 Announcement
15:30-15:40 ASC16 Top16 Team Video
15:40-15:50 ASC16 Top16 Team Speech
15:50-16:00 Closing Speech
 

Time

Agenda

16:00-16:30

ASC16 Final Competition Rules
Weiwei Wang, State Key Laboratory of HPC Server & Storage Technology

16:30-16:50

ASC16 Final Competition e-Prize Rules
Qing Zhang, State Key Laboratory of HPC Server & Storage Technology

16:50-17:10

Instructions of Tianhe-2 System
Chief Engineer of Tianhe-2 System Management

17:10-18:10

OpenACC Parallel Programming
Chandler Zhou, HPC Developer Engineer at NVIDIA

 

ASC16 Final Competition Notification.pdf

Contact Us

Tech Inquiry
Wenjing Lv & Weiwei Wang
TechSupport@asc-events.org

Hybrid Comp Inquiry
Qing Zhang
HybridComputing@asc-events.org

Media Inquiry
Jie He
Media@asc-events.org

Cooperation Inquiry
Xiaochen Fan
Partner@asc-events.org

General Inquiry
Yan Yan
Executive.Director@asc-events.org

Other Inquiry
Ying Zhang
Info@asc-events.org

Follow Us
Partners
Copyright 2016 Asia Supercomputer Community. All Rights Reserved