GPU Compiler Intern, OpenCL Compiler Team

AMD is looking for a spring intern to work with our core team developing OpenCL, an open standard for heterogonous general purpose programming, compilers for multi-core CPU and many-core graphics systems. This is a paid position.

The intern tasks usually range from implementing extensions to OpenCL that touch the entire compiler stack all the way down to device specific optimizations. As development will be with LLVM, knowledge of LLVM or contribution to LLVM a must. Knowledge of OpenCL API or other GPGPU programming models is a plus but not required. Experience with LLVM code generators and especially with MC infrastructure is a plus. The individual will be a member of a team where communication and team skills are highly valued.

Our intern this past summer worked on building the code generation foundation for our future FSA virtual ISA announced at AFDS`11.

The previous intern before that worked on memory vectorization optimizations in LLVM.

What do you want to contribute?

What we are looking for:

Basic knowledge of compiler implementation and design. Strong C and C++ programming skills is a must, along with an understanding of standard software engineering practices.

Exposure to algorithms used for optimizations and code generation for CPU or graphics hardware is a strong plus but not a must, as the primary sought quality is a strong interest and ability in developing high quality software on top of lower level functionality.

Experience with software development tools such as source level debugging, version control and code profiling is required.

If interested, please send resume to micah dot villmow at amd dot com. Sorry for the short notice, but deadlines for resumes is Friday November, 18th.


Micah Villmow