Sort by
Refine Your Search
-
Listed
-
Country
-
Field
-
infrastructures, improving system performance, scalability, and efficiency by optimizing resource usage (e.g., GPUs, CPUs, energy consumption). Researchers and students will explore innovative approaches to reduce
-
networking technologies such as InfiniBand. Working knowledge of GPU technologies like CUDA and OpenCL. Experience with distributed computing job schedulers (e.g., Slurm, PBS). Familiarity with
-
GHz with 24-cores each) and 4 GPUs (NVIDIA Ampere A100 80 GB PCIe) with 512 GB RAM. The compute nodes are interconnected via a fast InfiniBand network that also connects to ~360 TB of compute storage
-
and scalable model deployment. Skilled in working with medium-large scale multicore and heterogeneous (CPU + GPU) clusters. Excellent verbal and written communication skills. Preferred qualifications
-
Windows and Linux virtual machines for research and instructional purposes, including the allocation and optimization of GPU resources. Develop and maintain scripts and simple web applications to support
-
data analysis, simulation, and machine learning, integrating resources across multiple facilities. NERSC's next major supercomputer, Doudna, will combine next generation GPUs, networking and storage
-
well as experience working with large biological datasets in scalable GPU-based computing environments. What we provide: A competitive compensation package, with comprehensive health and welfare benefits. A supportive
-
GPU acceleration (CUDA) Participation in relevant competitions (e.g., Kaggle, computer vision challenges) Experience with version control (Git) and collaborative development practices
-
facilities. Core responsibilities include the deployment and maintenance of small-scale HPC and compute nodes, GPU workstations, Linux and Windows servers, research data storage and backup solutions
-
Implement sustainable and reproducible and FAIR research software engineering practices Collaborate with other HPC facilities and project partners Help evaluate and integrate GPU acceleration and other modern