Differences

This shows you the differences between two versions of the page.

--- opt:graphics_cards [2018/08/31 08:56]
pklapetek
+++ opt:graphics_cards [2018/08/31 09:52]
pklapetek
@@ Line 53: / Line 53: @@
 <html>
-GPU
+<tt>
+GPU<br>
+<br>
-UGPU
+<br>
+UGPU<br>
-<html>
+<br>
+</tt>
+</html>
 will run the calculation on the Tesla K40m card on our system.
+Even if we try to do our best to have one to one correspondence for everything on CPU and GPU, it can happen that some of the recent algorithms are not yet implemented on GPU. The reference should always be the CPU implementation. We therefore strongly recommend, namely for setting up the new task, to test the task first on CPU, crosscheck it with GPU and then run the various repetitive calculations on GPU to speedup the solution.
+As an example you can test the following parameter files, comparing performance of GSvit on different number of CPU cores and on a single GPU on the above mentioned system. The calculation took 8 minutes 37 seconds on a single core, 1 minute 43 seconds on eight cores, and 54 seconds on the first GPU.
+----
+//
+{{ :samples:img_gpu.png?100|}}
+Sample parameter file: {{samples:gpu.tar.gz|gpu use}}.
+\\
+A 200x200x200 computational domain with a plane wave and dielectric sphere, run on CPU, CPU with multithreading and on GPU. Three different parameter files are provided, one for each case.
+//
+----
+Generally, the bigger is the calculation size and problem complexity, the better GPU works comparing to CPU as the effect of administration (data transfers) is smaller. Also, too many unnecessary transfers, e.g. for image outputs at every few steps, degrade the GPU performance.

GSvit documentation

User Tools

Site Tools

Differences

Page Tools