The Intel C++ Compiler is a full fledged C/ C++ compiler and debugger suite. This 8.1 release features support for the new Intel Extended Memory 64 technology and integration with Eclipse IDE.
Intel C++ Compiler 8.1 Released
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
What is that integration?
Thanks.
…from Fortran 95, “What am I, chopped liver?” To wit:
http://developer.intel.com/software/products/compilers/flin/index.h…
Yes, IFC 8.1 was also released. Now you can compile your real code for Opterons. The people of Molpro at Cardiff are showing IFC 8.1 + ATLAS as being a great combo for Opterons.
That said, still not a lot of benchmarks for 8.1 + EM64T, which I’m hoping this’ll spur.
Yes, IFC 8.1 was also released. Now you can compile your real code for Opterons.
Haven’t heard of http://www.pathscale.com/ ?
OK, I have a question. Why is it that IFC (and ICC?), don’t have a “host”-type optimization flag? That is, on IFC, I can use -xN or -xP to get P4 optimizations, and that’s great. But, you have to remember to change to -xK, say, if you move your code to a PIII.
With ye olde Compaq Fortran for my Alpha, though, I can use -arch host or -tune host and it’ll pick out whether I have an ev6 or ev67 and compile appropriately.
So, is there something about the x86 architecture that Intel can’t have a -xQ, say, where it’ll probe for chip type and select the appropriate -x{} option?
Jus’ wondering…
I do know Pathscale, and it compiles code that is slower than IFC 8.1 and PGF 5.2 on benchmarks I’ve seen, given the same BLAS/LAPACK library. Plus, isn’t Pathscale (barring promotions) more expensive than IFC?
How it would behave on an AMD64…
I think a good number of people are forgetting SUN’s compiler, which IIRC will be Opteron “ready” for Solaris for x86-64 once released. If their compiler is as good as their SPARC version, Opteron will have a great opportunity to have its power fully exploited.
wasn’t there an article on here that showed the Sun compiler not to be all that hot? (not trolling, I’m a Sun fan)
I think expecially within the strongly competitive market as that of IA-32/64/AMD64, Sun’s pricing (US$ 3000 per license) ist just prohibitive – or quite natural that tey are fogotten. There are many highly specialized HPC compiler manufacturers with good reputations (PGI, Absoft, …) out there, with compilers that are still 4 digits, but noticably less expensive.
I think this is the article you’re referring to. http://www.osnews.com/story.php?news_id=5830&page=1
Sun’s compiler doesn’t seem to be much better than GCC in these benchmarks.