Loading...
You are not logged-in Login/Register





  • Posts   Search Threads
  • Clay Breshears (Intel)July 2, 2009 10:41 AM PDT   
    Scoring Criteria

    Four test cases were used to score execution time.

    Case 1: 10 e.Coli segments queried with the same data parsed into 800 base segments
    Case 2: 400 e.Coli segments queried with the 800 base segments from 10 e.Coli segments
    Case 3: 12 one million base segments quereied with 10 e.Coli segments
    Case 4: 12 one million base segments queried with 400 e.Coli segments

    One addition to the query file for the first two cases was a string of "TTTTTT".

    Case 1 should find all query strings at regular 800 base intervals. Case 2 should have the same output as Case 1 with extra segment matches in segments 11-400. Case 3 should have no matches for any of the 10 queries, and Case 4 should have four total segment matches (repeated 3 times due to repeated data in data base file). A maximum of two minutes was set for each input test case.

    There were 10 submissions of code solutions, 5 on Linux and 5 on Windows. There was one entry written in C#, one threaded with Cilk++, and the rest in C/C++. Two entries had runtime issues or errors (but not for all test cases), and one entry time out in all test cases (after sucking up 98.7% of the available memory). Penalty points were assessed to entries for coding problems and incomplete output.

    Point spread:

    100
    99 99 99
    89 89
    49
    25
    19
    0(x4)

    The write-up portion of each entry was read and scored by two judges. Each judge used the 10-30-10 breakdown of points for serial algorithm description, parallel algorithm description, and performance, respectively. One important component to the judging was to determine how close the submission was for publication on ISN. The assigned score was the average of the two judges scores.

    Point spread:

    44
    29 27 22 21
    17 13 10
    8 6 5
    0 0

    Bonus points were given for contestant’s forum posts made before the problem entries were closed. Five points per post (maximum 25 points possible) were awarded.

    The overall winner was Bradley Kuszmaul, who also had the highest scoring write-up submission. The fastest code execution was submitted by akki.

    --clay

    P.S. I appreciate the patience of all the contestants for this problem and the last two of the series with regards to my absence and the delays that has caused with scoring and announcing winners.  (I hope to take the intervening months and get my act together with some planning and other prep work to have the next phase run smoother and faster.)

    Sergii BiloshytskyiJuly 2, 2009 11:51 AM PDT
    Rate
     
    Re: Scoring Criteria

    May contestants have any feedback on their submissions? At list a scoreboard or something.. For example, it would be nice for me to know how my program behaved during tests.

    Thanks.


    matteocilk.comJuly 3, 2009 5:41 AM PDT
    Rate
     
    Re: Scoring Criteria

    Congratulations Bradley!


    mikhailsemenovJuly 3, 2009 5:43 AM PDT
    Rate
     
    Re: Scoring Criteria

    Four test cases were used to score execution time.

    Case 1: 10 e.Coli segments queried with the same data parsed into 800 base segments
    Case 2: 400 e.Coli segments queried with the 800 base segments from 10 e.Coli segments
    Case 3: 12 one million base segments quereied with 10 e.Coli segments
    Case 4: 12 one million base segments queried with 400 e.Coli segments

    One addition to the query file for the first two cases was a string of "TTTTTT".

    Case 1 should find all query strings at regular 800 base intervals. Case 2 should have the same output as Case 1 with extra segment matches in segments 11-400. Case 3 should have no matches for any of the 10 queries, and Case 4 should have four total segment matches (repeated 3 times due to repeated data in data base file). A maximum of two minutes was set for each input test case.

    There were 10 submissions of code solutions, 5 on Linux and 5 on Windows. There was one entry written in C#, one threaded with Cilk++, and the rest in C/C++. Two entries had runtime issues or errors (but not for all test cases), and one entry time out in all test cases (after sucking up 98.7% of the available memory). Penalty points were assessed to entries for coding problems and incomplete output.

    Point spread:

    100
    99 99 99
    89 89
    49
    25
    19
    0(x4)

    The write-up portion of each entry was read and scored by two judges. Each judge used the 10-30-10 breakdown of points for serial algorithm description, parallel algorithm description, and performance, respectively. One important component to the judging was to determine how close the submission was for publication on ISN. The assigned score was the average of the two judges scores.

    Point spread:

    44
    29 27 22 21
    17 13 10
    8 6 5
    0 0

    Bonus points were given for contestant’s forum posts made before the problem entries were closed. Five points per post (maximum 25 points possible) were awarded.

    The overall winner was Bradley Kuszmaul, who also had the highest scoring write-up submission. The fastest code execution was submitted by akki.

    --clay

    P.S. I appreciate the patience of all the contestants for this problem and the last two of the series with regards to my absence and the delays that has caused with scoring and announcing winners.  (I hope to take the intervening months and get my act together with some planning and other prep work to have the next phase run smoother and faster.)

    I guess I did not get anywhere because I had not attached the files that I had downloaded.

    邓辉July 3, 2009 6:02 AM PDT
    Rate
     
    Re: Scoring Criteria

    Congratulations Bradley! :)


    写字楼里写字间,写字间里程序员
    程序人员写程序,又拿程序换酒钱
    酒醒只在网上坐,酒醉还来网下眠
    酒醉酒醒日复日,网上网下年复年

    mikhailsemenovJuly 3, 2009 6:27 AM PDT
    Rate
     
    Re: Scoring Criteria

    Congratulations Bradley! Best regards to akki!


    akkiJuly 3, 2009 8:23 AM PDT
    Rate
     
    Re: Scoring Criteria

    Quoting - mikhailsemenov
    Congratulations Bradley! Best regards to akki!

    Fastest code! Cool!

    Thanks mikhailsemenov


    akkiJuly 3, 2009 8:41 AM PDT
    Rate
     
    Re: Scoring Criteria

    Quoting - akki

    Fastest code! Cool!

    Thanks mikhailsemenov

    I've updated my unofficial scoreboard. Here's the link:
    http://spreadsheets.google.com/ccc?key=rRPX2nM92vYSLUPDmjQgF7A


    haojnJuly 4, 2009 12:00 AM PDT
    Rate
     
    Re: Scoring Criteria

    Congratulations Bradley!


    haojnJuly 4, 2009 12:05 AM PDT
    Rate
     
    Re: Scoring Criteria

    Quoting - akki

    I've updated my unofficial scoreboard. Here's the link:
    http://spreadsheets.google.com/ccc?key=rRPX2nM92vYSLUPDmjQgF7A

    Thank you!

Forum jump:  

Intel Software Network Forums Statistics

17,025 users have contributed to 48,317 threads and 172,754 posts to date.

In the past 24 hours, we have 9 new thread(s) 56 new posts(s), and 52 new user(s).

In the past 3 days, the most popular thread for everyone has been How to manage rounding by IVF ?? The most posts were made to Most likely, the issue is that The post with the most views is Optimalization of sine function\'s taylor expansion

Please welcome our newest member redfruit83


For more complete information about compiler optimizations, see our Optimization Notice.