#openmp

Patrick Ian Smith | 22 years old | Manic Depressive, Drug Addiction | FC: Ezra Miller | OPEN

If one thing is certain, Patrick always loved music. This passion paid off when at the age of 19, he was skyrocketed into stardom. Thus began his mental breakdown. Patrick had been diagnosed bipolar but didn’t let it get to him. He quickly became a sort of teen heart throb, and eventually married a supermodel at twenty. Patrick was, on the surface, happy. But, underneath the glamor of rock and roll was a beast just waiting to rear its ugly head. One night, the opportunity arrived. Patrick had been under the influence of several drugs as well as copious amounts of stress. He became completely delusional, believing that he was being watched by the government. He destroyed seven mailboxes, two squad cars, and a 2.5 million dollar statue. It was then that his wife decided she could no longer put up with his constant mood swings, and shipped him off to Pineridge with divorce papers.

GPUにプログラムを投げるならどれがいいの

Sam Oliver Bosner | 19 years old | Depressive, Alcoholic | FC: Cole Sprouse | OPEN

Sam never knew his mother. She’d been anorexic, and severely underweight when she’d gone into labor. Sam was her fourth child, all other’s having been girls. Due to complications during childbirth, and having been severely malnourished, Angela died and Sam was born completely deaf. Due to his lack of hearing, angsty birth and being the youngest, his sisters absolutely coddled him. He learned to write in order to communicate with the general public, keeping a note pad handy at all times. Sam can only speak sign language. His feelings of isolation led him to become depressed. Sam took to the bottle as a means of coping with his guilt over the death of his mother. No longer being able to keep up with his son, Mr. Bosner sent him off to Pineridge.

Theo Helford | 23 years old | Insomnia, OCD, Super-sensitivity | Face Claim: Donald Glover | OPEN

Theo was diagnosed with supersensitivity at a young age but was expected to grow out of it. He was a bright kid with a bright future: he got fantastic grades and blew everyone away with his sports skills. He earned a sports scholarship and aced his exams – he was set for life. Gradually, he started to develop OCD, related to his supersensitivity. At first, it was simple enough. Different foods couldn’t touch; his clothes couldn’t be made of certain fabrics, etc. Then, over time, it stemmed into more general, anxiety-based OCD. Washing his hands twice after every meal, folding all his clothes and especially that everything had to be in its place. But still, he managed, and so did his family. 

Working on an OpenMP application I came across an interesting race condition. The application involved iteration over the contents of a matrix, using a for-loop and a second nested for-loop, during which the values in each row were arithmetically modified by an offset proportional to the foremost non-zero value in each column. Within the serial implementation, in the first for-loop, the offset was computed, prior to modification of the values within a given row using the computed offset.

A naive approach to the optimization of the application was to introduce a parallel section within the outer for-loop to perform the arithmetic modification of each value within a given row of the matrix in parallel. This was done by simply by introducing a ‘pragma omp parallel’ section around the contents of the outer for-loop, and a ‘pragma omp for’ section around the inner for-loop.

After running the application and observing that the production of the expected output was inconsistent, I hypothesized that there was likely a race condition in the application. After spending time debugging, I observed that, since the offset was proportional to the foremost, non-zero value in the current row, and that a thread would modify this value, there was a race condition which necessitated the synchronization of the threads in the parallel section after the computation of the offset, in order to ensure that all threads had computed the same offset before proceeding to modify the contents of the matrix using that offset, since one of the threads, in modifying the matrix, would alter subsequent computations of the offset. After the addition of a ‘pragma omp barrier’ which followed the computation of the offset in the parallel section, before the inner for-loop, the application consistently produced the correct result.

前回の記事で、elemental関数をいろいろと取回せるようになりました。

どうせなので、OpenMP並列化をしてみようと思います。

あと、副作用のあるelement関数についても触れてみます。