non commercial license renewing

[benlala_s_]

Dear mumbers,

I want to renew my non-commercial license for Intel Parallel Studio XE for Fortran and C++ for linux.

How I do this ?

[Steven_L_Intel1]

Sorry, we have discontinued non-commercial licensing.

[FlyingHermes]

May I ask why ?

[TimP]

Did you read the comments elsewhere about widespread violations of non-commercial license terms, particularly by those who qualify for academic licensing?

[Roc_W_]

Let's be honest, I have to agree with Intel's observation that many academic users had been using their non-commercial offerings. However I somehow felt Intel's reaction wouldn't really change the situation but merely make it inconvenient. As I have observed, many research groups don't really have much funding for this type of software purchase, even if they write some form of in-house code for their projects. Groups that are supported to do major software development as one of their project goals do purchase proper license. So with discontinuance of the non-commercial offerings, I suspect one of a few things may happen: (1) sticking to the old version of Intel compilers one has on his workstations; (2) moving the development to supercomputing centers (where the production runs will be performed and where they do always have the license), suffering occasional problems with slow/remote connections; (3) using a different compiler like gfortran, sunf95, pathf95, openf95 locally, and maybe, changing to Intel on the production machines. In this case one might get locked in to other compilers by their non-standard language extensions (as happened to me with Intel). These are of course just my experiences/observations and speculations. Intel will have the data to tell what really will happen. Frankly I think none of these will make its current license holders deflect, as long as Intel CPUs continue to sell and its compilers produce faster executables; they just might to some degree reduce the good will Intel currently has with its users (and not increasing its sales much). It's frustrating especially considering that developers of many open-source programs have contributed a lot to discovering bugs in their products, both C/C++ and Fortran..

[Steven_L_Intel1]

From what the business people tell me, the bigger problem was commercial use of non-commercial licenses. I agree with your thoughts on what the effect of this will be for academic users, though.

[Roc_W_]

Oh OK. I thought the move is targeting mainly at academic users, many of which use Fortran, and the non-commercial C/C++ compiler would return. Yeah, there is no way that the majority kind of research groups that I know of would have the appropriate budget to buy compilers for students' workstations. They will just be asked to use a free alternative, do things remotely on their computing center, use cross-compiling if people absolutely want to use Intel for their local machines...

[Steven_L_Intel1]

We have VERY deep discounts for students and classroom packs.

[Miguel_H_]

It is true, that very deep discounts are available for students, and indeed at $129 it is definitely a great deal.

However, the problem I see is with MS and PhD students, which are the ones that are most often going to use the compilers for their dissertations. Virtually all of them get a fellowship, meaning that they are being paid for doing their research, and this makes them not eligible for the student price.

And the academic license for more than one user is just too expensive for many research groups, especially if the national research funding programs do not give money for buying software, which is for example the case in my country (we can argue if it makes sense or not, but it is like that). At the end the same will happen as with other software packages, namely that one license is bought (if at all) by the Professor for himself in order to get the support and be able to submit bug reports, and its students will get their copies of the compilers through the "dark corners" in the internet.

[Jeyan_T_]

As others have already pointed out, research groups in universities do not have much funding to invest on compilers --- even with deep discounts. The violation issues are understandable- but I very much disagree that withdrawal is the only option. We will be left with no option but to steer towards free versions (and to motivate our PhD students / researchers to do the same). The net result is that they will, in turn, influence their future employers to do the same!.

[Edgardo_Doerner]

As Academic I totally agree with the comments above. In my case I purchased the academic license for myself, but I must always be careful to be able to run my codes with the GCC suite in order to have the "free choice" always available for the students. It is really a shame, considering the importance that the Intel compilers with its valuable debugging and optimization tools have had in my research, but I cannot ask my students to buy a license, specially when they are not eligible to do so.