<div dir="ltr"><div><div>Since we use our cooked version of parallel Gate, the ROOT output file </div>is split in hundred of files, and they usually do not reach a big size. But this </div>is not a real solution to the ROOT problem. Fernando </div><div class="gmail_extra"> <div class="gmail_quote">On Mon, Mar 3, 2014 at 11:49 PM, 张春晖 <<a href="mailto:arnoxiaolei86@gmail.com" target="_blank">arnoxiaolei86@gmail.com</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hello Fernando I am indeed using ROOT as the output now. How did you resolve it? Thanks a lot! Best!--Chunhui Zhang 2014-03-03 23:54 GMT+08:00, Fernando Rannou <<a href="mailto:fernando.rannou@usach.cl">fernando.rannou@usach.cl</a>>: > I have experienced that if I use ROOT for data recording, the I/O gets > slower and slower if the file > grows excessively. > > > Fernando > > > On Mon, Mar 3, 2014 at 11:45 AM, 张春晖 <<a href="mailto:arnoxiaolei86@gmail.com">arnoxiaolei86@gmail.com</a>> wrote: > >> Hello Simon: >> Thank you for your reply! When simulating our PET system, I >> found that my Gate became slower and slower, Have you encountered >> it？In addition, What is the meaning of “parallelizing our >> simulation”？ using the Cuda? or dividing our simulation into many >> small tasks? >> >> Looking forward to your reply, Thanks a lot! >> Best.---Chunhui Zhang >> >> >> 2014-03-03 16:51 GMT+08:00, Simon Stute <<a href="mailto:gate.stute@gmail.com">gate.stute@gmail.com</a>>: >> > Hi, >> > >> > Gate is build on Geant4 in a quite straightforward way. >> > So to answer your question "is G4 faster than Gate", I would say: "it >> > depends on how you will code your new program". >> > G4 cannot be used on its own, it is a library. >> > If you want a really really simple PET simulation, maybe you can do it >> with >> > G4. >> > But if you want a "real" PET simulation in the "realistic" sense, then >> > I >> > recommend you to stay on Gate. >> > Otherwise, you need good skills in C++ and plenty of time to do >> > something >> > already done... >> > Also you won't benefit from a validated and experienced code. >> > At least, have you thought about parallelizing your simulation ? >> > >> > Regards, >> > Simon >> > >> > >> > >> > On Mon, Mar 3, 2014 at 4:09 AM, 张春晖 <<a href="mailto:arnoxiaolei86@gmail.com">arnoxiaolei86@gmail.com</a>> wrote: >> > >> >> Hello everyone: >> >> I am simulating the small animal PET using Gate V6.1， but the >> >> speed is so slow. I want to know whether the speed of Geant4 is >> >> faster than that of the Gate. If so, I will simulate our system under >> >> Geant4, Thanks a lot! ---Chunhui Zhang >> >> _______________________________________________ >> >> Gate-users mailing list >> >> <a href="mailto:Gate-users@lists.opengatecollaboration.org">Gate-users@lists.opengatecollaboration.org</a> >> >> <a href="http://lists.opengatecollaboration.org/mailman/listinfo/gate-users" target="_blank">http://lists.opengatecollaboration.org/mailman/listinfo/gate-users</a> >> > >> _______________________________________________ >> Gate-users mailing list >> <a href="mailto:Gate-users@lists.opengatecollaboration.org">Gate-users@lists.opengatecollaboration.org</a> >> <a href="http://lists.opengatecollaboration.org/mailman/listinfo/gate-users" target="_blank">http://lists.opengatecollaboration.org/mailman/listinfo/gate-users</a> > </blockquote></div> </div>