Employer Active
Job Alert
You will be updated with latest job alerts via emailJob Alert
You will be updated with latest job alerts via emailACG2643JOB
Our client a leading technology company is looking for a qualified candidate to join their firm.
Implement and enhance different artificial neural network layers and tensor operations using technologies like OpenCL and CUDA.
Develop models for predicting performance and analyzing tensor operations across various hardware accelerators.
Create code generators and autotuners to produce and select the best kernel code tailored for diverse GPU architectures.
Make use of GPU components designed for AI tasks including dedicated matrix multiplication units and highspeed GPU interconnections.
Design efficient communication libraries to facilitate GPUtoGPU communication including interactions between GPUs across different nodes.
Bachelors degree in Computer Science or related field with an exellent GPA and/or excellent academic performance.
Proficiency in C and software development.
Experience with OpenCL or CUDA development and a solid understanding of GPU architecture.
Indepth understanding of computer systems.
Proficient in English.
Preference
Masters degree
Extraordinarily brilliant individuals (e.g. IMO/IOI Medalists ICPC National Olympiad in Informatics/Math etc.
Preferential support for professional research personnel
Experience in parallelizing/optimizing AI/HPC applications on GPUs with a history of related paper publications
Experience developing programs for GPU clusters using MPI NCCL etc.
Participation in largescale software development based on C
Basic knowledge of deep learning and experience implementing artificial neural network models
History of paper submissions to major academic conferences
Fulltime salary during training
Training and cultural experience in Korea
Fully covered expenses: Visa fees transportation meals and accommodation provided
Opportunities to join international projects and tasks
Study and practice at University guidance from experts
Contact: Thao Phan
Due to the immense number of applications only shortlisted candidates will be contacted.
Full Time