1 month ago
Taiwan (city not specified)Staff+
Responsibilities
- Design AI HPC architectures that optimize performance, scalability, and efficiency for AI workloads.
- Evaluate and select hardware, networking, storage, GPU acceleration, and interconnect resources for large-scale AI computation.
- Optimize AI algorithms and workloads using parallel processing, distributed computing, and GPU acceleration.
- Tune system configurations, resource allocation, and workload management while resolving computation, memory, storage, and network bottlenecks.
- Design scalable, highly available, fault-tolerant systems using redundancy, load balancing, and failover mechanisms.
- Collaborate with software engineers and DevOps teams to integrate AI models into production and develop deployment workflows.
- Evaluate emerging AI, machine learning, and HPC technologies, frameworks, and tools.
- Document system designs and optimizations and communicate technical recommendations to diverse stakeholders.
Requirements
- Deep understanding of artificial intelligence, machine learning concepts, algorithms, and frameworks, with experience developing and deploying AI models.
- In-depth knowledge of high-performance computing architectures, parallel processing, distributed computing, GPU acceleration, and HPC best practices.
- Experience designing scalable system architectures and configuring hardware, networking, and storage for high-performance computing.
- Proficiency in Python, C++, or Java; Bash or PowerShell scripting knowledge is beneficial.
- Familiarity with MPI, OpenMP, CUDA, or OpenCL for parallel computing and GPU acceleration.
- Experience with performance tuning and optimization for large-scale computing systems, including profiling, benchmarking, and workload management.
- Understanding of system administration, Linux or UNIX-based operating systems, system monitoring tools, and cluster management frameworks.
- Strong communication and collaboration skills with data scientists, researchers, software engineers, and IT professionals.
- Commitment to continuous learning and staying current with AI and HPC research and technologies.
About TSMC
TSMC (Taiwan Semiconductor Manufacturing Company) is a pure-play semiconductor foundry that manufactures integrated circuits for fabless and systems companies on a contract basis. Founded in 1987 and headquartered in Hsinchu, Taiwan, it is the world’s largest chip foundry, with advanced process nodes and fabs in Taiwan and new facilities in Arizona and Japan. Its shares trade on the Taiwan Stock Exchange (2330) with ADRs on the NYSE (TSM).