I’m not quite sure this post belongs here, but I couldn’t find another appropriate channel, so I was hoping one of the nVIDIA employees around here might point me in the right direction. My question is as follows: I am a postdoctoral research scientist using CUDA to accelerate my institution’s data reduction techniques. As required, I am preparing journal papers on my work for publication, and I would like to know who to contact to obtain permission to reproduce/include a couple figures from the programming guide in an appendix of one of my papers, and also if there is a standard acknowledgment statement required by nVIDIA for published works using CUDA on their GPUs.
A pointer to an appropriate representative would be most appreciated. Thank you!
Just give a reference to the document from which you’re obtaining the information, the way you would for any other source.
It would be cool if NVIDIA could provide “the usual figures” from the progguide and the best practice guide in a separate download: “grid of blocks of threads” etc. I stick to not including them in my pubs because all I can get is a screenshot, which most journals do not accept :) On the other hand, since there are more than 1000 papers out on CUDA already, there is no point in reiterating the programming model over and over again.
You’d think wouldn’t you? Sadly, I still see reviews making statements and asking questions that pertain to the programming model. I’ve seen papers rejected simply because they didn’t have this information and the reader was confused because he/she was not familiar with the programming model/architecture. Publishing… like I’ve said.
Thank you all for the suggestions. Paulius’s suggestion is probably what I’ll end up doing. I was concerned at first, because citing a publication from a referreed journal is standard practice, but I figured the protocol would be different, since I’d be citing resources from a private corporation. Thanks guys!