Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <h2>Synchronization overhead</h2> <p>I would guess that <strong>the amount of work done per iteration of the loop is too small</strong>. Had you split the image into four parts and ran the computation in parallel, you would have noticed a large gain. <strong>Try to design the loop in a way that would case less iterations and more work per iteration</strong>. The reasoning behind this is that there is too much synchronization done.</p> <h2>Cache usage</h2> <p>An important factor may be how the data is split (partitioned) for the processing. <strong>If the proceessed rows are separated</strong> as in the bad case below, then <strong>more rows will cause a cache miss</strong>. This effect will become more important with each additional thread, because the distance between rows will be greater. If you are certain that the parallelizing function performs reasonable partitioning, then manual work-splitting will not give any results</p> <pre><code> bad good ****** t1 ****** t1 ****** t2 ****** t1 ****** t1 ****** t1 ****** t2 ****** t1 ****** t1 ****** t2 ****** t2 ****** t2 ****** t1 ****** t2 ****** t2 ****** t2 </code></pre> <p>Also make sure that you <strong>access your data in the same way it is aligned</strong>; it is possible that each call to <code>offset[]</code> and <code>BayerChannel[]</code> is a cache miss. Your algorithm is very memory intensive. Almost all operations are either accessing a memory segment or writing to it. Preventing cache misses and minimizing memory access is crucial.</p> <h2>Code optimizations</h2> <p>the optimizations shown below may be done by the compiler and may not give better results. It is worth knowing that they can be done.</p> <pre><code> // is the memset really necessary? //memset(RgbChannel, 0, Width * Height * 3 * sizeof(T)); parallel_for(0, Height, [&amp;] (int row) { int rowMod = (row &amp; 1) &lt;&lt; 1; for (auto col = 0, bayerIndex = row * Width, tripleBayerIndex=row*Width*3; col &lt; Width; col+=2, bayerIndex+=2, tripleBayerIndex+=6) { auto rgbIndex = tripleBayerIndex + offsets[rowMod]; RgbChannel[rgbIndex] = BayerChannel[bayerIndex]; //unrolled the loop to save col &amp; 1 operation rgbIndex = tripleBayerIndex + 3 + offsets[rowMod+1]; RgbChannel[rgbIndex] = BayerChannel[bayerIndex+1]; } }); </code></pre>
    singulars
    1. This table or related slice is empty.
    1. POParallelizing a for loop gives no performance gain
      singulars
      1. PTQuestion
    1. PTAnswer
    1. USDariusz
    1. USDariusz
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    1. CO"too much synchronization done" - I think that Intel TBB's [`parallel_for`](http://threadingbuildingblocks.org/docs/help/reference/algorithms/parallel_for_func.htm) is smart enough to do right partitioning: "However, for efficiency, do expect parallel_for to tend towards operating on consecutive runs of values"
      singulars
      1. PO
      1. USEvgeny Panasyuk
    2. COI don't know about concurrent::parallel_for but TBB doesn't create a thread per iteration but rather partitions the job into the number of "available cores". At least that's the way it was when I used it with Intel compiler 11
      singulars
      1. PO
      1. USeladidan
    3. CO@eladidan: MS PPL is almost the same thing as Intel TBB
      singulars
      1. PO
      1. USAndriy Tylychko
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload