AMD Releases Optimizing C/C++ Compiler For Ryzen

[Michael]

Michael , are you planning on some Ryzen benchmarks with llvm vs aooc? Would be particularly interested in applications that use AMD's AVX implementation

As mentioned in the article, yes.

[dungeon]

RESTRICTIONS. Except for the limited license expressly granted in Section 2 herein, You have no other rights in the Software, whether express, implied, arising by estoppel or otherwise. Further restrictions regarding Your use of the Software are set forth below. You may not:


modify or create derivative works of the Software;
distribute, publish, display, sublicense, assign or otherwise transfer the Software;
decompile, reverse engineer, disassemble or otherwise reduce the Software to a human-perceivable form (except as allowed by applicable law);
d) alter or remove any copyright, trademark or patent notice(s) in the Software; or
e) use the Software to: (i) develop inventions directly derived from Confidential Information to seek patent protection; (ii) assist in the analysis of Your patents and patent applications; or (iii) modify existing patents.
f) use the Software in a way that requires that the Software or any portion thereof be licensed under a Free Software License.

That one is pretty interesting, fuck you AMD

[sdack]

Judging only by the description given in the article does it sound very much like they couldn't have cared less if they've tried and that this was the last compiler ever to have come from AMD.

I'm curious if the benchmark will reflect this or if it turns out to be some wonder-magic compiler against all expectations.

[ossuser]

Seeing AMD going opensource big time, I don't understand this.

Does this also mean that gcc /clang didn't get the full set of Ryzen optimizations yet ?

Might this be a 'temporary' compiler for demonstrating the power of Ryzen 7/Whitehaven/Naples etc ?

@bridgman: any idea ?

[ermo]

But the application has to be compiled with a specialized architecture flag? Thus, no one will benefit from it?

Erh ... surely people who turn on the relevant architecture flag will benefit?

Or are you assuming that the people who want the extra performance won't recompile from source for some unspecified reason?

[bridgman]

yeahhh I don't think this is gonna be upstreamed

That one is pretty interesting, fuck you AMD

Huh ? The compiler teams have been submitting Ryzen support upstream into LLVM for quite a while now.

This is just a binary release with the standard binary EULA as far as I know.

Why would we try to push a binary upstream ?

[zekesonxx]

Why would we try to push a binary upstream ?

Because it sure looks like you took clang, made some Ryzen optimizations, and are now trying to make your own icc/icl.

While technically legal, it's a shitty thing to do. It seems like a very backwards step given all of your recent open-source work.

[bridgman]

What part of "we have been pushing changes upstream for a quite a while" was not clear ?

[Steffo]

I won't use any vendor compiler!

[juno]

Why would we try to push a binary upstream ?

Why would marketing claim it is an "AMD compiler" if it's just upstream llvm instead of advertising that the company is contributing to free software?