<p dir="ltr">Hello,</p>
<p dir="ltr">I still haven't solved the problem. <br>
Could anybody please confirm that the GPU example (examples/example_GPU/pet) is working properly after renaming 'thorax_phantom.bin' to 'throrax_phantom.i33'?</p>
<p dir="ltr">In my case there is a huge difference between the CPU and GPU simulation output that I can't explain... Any ideas?<br></p>
<p dir="ltr">Best, </p>
<p dir="ltr">Mirco</p>
<div class="gmail_quot<blockquote class=" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div>Dear Gate users,<br><br></div>I have successfully installed Gate 7.0 and the Nvidia CUDA tools and am now trying to run the GPU example.<br>However, it seems that when using the GPU the coincidence sources are not located inside
the FOV but mainly along the faces of the image volume. When running the same simulation using the CPU simulation file everything seems to be fine, as visualized in the first picture.<br></div><div><br>Do I have to change more than the following lines? It was mentioned before that there might be an issue with the setPosition command (<a href="http://permalink.gmane.org/gmane.comp.science.opengate.user/3929" target="_blank">http://permalink.gmane.org/gmane.comp.science.opengate.user/3929</a>). I have already set it to the position of the CPU example and the origin without resulting difference.<br><br></div><div># CPU<br></div><div>/gate/source/addSource src voxel<br>...<br>/gate/source/src/setPosition -25.6 -12.6 -9.2 cm<br><br><br></div><div># GPU<br></div><div>/gate/source/addSource srcvoxel GPUEmisTomo<br>/gate/source/srcvoxel/attachPhantomTo ncat<br>/gate/source/srcvoxel/setGPUBufferSize 1000000<br>/gate/source/srcvoxel/setGPUDeviceID 1<br>...<br>/gate/source/srcvoxel/setPosition 0 0 0 cm<br><br><br></div><div>I have also tried to run a simulation using my own scanner and interfile data. The result was the same, meaning that the coincidence sources were mainly located on the faces of the image volume.<br><br><br></div><div>Any help is greatly appreciated.<br>Thanks a lot in advance and best regards,<br>
<br>
<div>
Mirco</div></div></div><br clear="all"><br>-- <br><div><div dir="ltr"><font size="-1"><b>Mirco Heß</b><br>
PhD Student<br>
European Institute for Molecular Imaging (EIMI)<br>
<p><img src="cid:part1.06060103.07070901@uni-muenster.de" alt="_European Institute for Molecular Imaging - EIMI_" height="70"></p></font><div><font size="-1">European Institute for Molecular Imaging - EIMI<br>
Waldeyerstr. 15<br>
48149 Münster<br>
<a href="http://www.uni-muenster.de/eimi" target="_blank">http://www.uni-muenster.de/eimi</a><br>
<br>
</font><font size="-1"><span><a href="tel:%2B49%20%280%29251%2083%20-%2049318" value="+492518349318" target="_blank">+49 (0)251 83 - 49318</a></span><br>
<a href="tel:%2B49%20%280%29251%2083%20-%2049313" value="+492518349313" target="_blank">+49 (0)251 83 - 49313</a> (Fax)<br>
eMail: <a href="mailto:mirco.hess@uni-muenster.de" target="_blank">mirco.hess@uni-muenster.de</a></font></div></div></div></div>
</div>