IBM is looking for an outstanding compiler engineer to work with a team developing the next generation of computational solvers for computational neuroscience and brain simulation targeting multiple parallel computing platforms including GPGPUs.
The project requires a skill set of developing optimizing compilers and designing runtime systems for parallel architectures. Of particular importance is source-to-source translation, schedule optimization. The successful candidate must have hands-on experience with compiler development. Experience with compiling openACC/openMP pragmas is definitely useful as well as knowledge of garbage collection algorithms.
It is a plus if the candidate has experience with coding in CUDA for GPGPUs, scripting languages (e.g. Python) and UNIX/LINUX tools.
The position requires daily interactions with group team members and with external collaborators, so good oral and written communication skills are preferred.
A little about us:
IBM is the world’s largest information technology company with more than 360,000 employees serving clients in 170 countries.