Infrastructure Solutions Architect, GPU

We are seeking a Singtel's GPU-as-a-Service (GPUaaS) Solutions Architect to assist in designing and implementing scalable and secure solutions that align with business objectives and technology standards. The incumbent will develop expertise in designing scalable and secure solutions that meet business requirements and technology standards in the team.

Responsibilities

• Analyze and translate business requirements into comprehensive technical solutions that align with Singtel's GPU-as-a-Service (GPUaaS).
• Design and implement architecture Singtel's GPU-as-a-Service (GPUaaS) based on business needs and industry best practices.
• Lead technical workshops and discussions to align architecture vision with business goals
• Co-design and develop applications/workload architecture with customers and partners and facilitate customer POC, trial and production requirement.
• Collaborate with business stakeholders, product managers, and technical teams to align on solution objectives and priorities.
• Manage relationships and expectations with key stakeholders and clients in the AI domain.
• Ensure architecture solutions are optimized for performance and scalability in GPU cloud environments and guide best practice architecture for Singtel's GPU-as-a-Service (GPUaaS).
• Ensure compliance with security, performance, and regulatory standards throughout the solution lifecycle.
• Lead on the planning and execution of the customer workload and data migration on our GPU platform.
• This role may require availability outside standard work hours, including nights, weekends and public holidays.

Requirements

• Degree in Computer Science, Information Systems, Engineering, or a related field with 8-10 years experiences.
• Strong hand-on experience in Linux, hypervisor, storage (NFS, Object), and infrastructure as a code.
• Understanding of cloud architectures (IaaS, PaaS), GPU system architecture and NVIDIA GPUs.
• A background in systems level thinking and design. Combined with the ability to translate technical strategy and architectures into concrete, minimal viable products.
• Experience architecting, designing, and developing complex, enterprise grade, configurable, scalable and high-performance architectures across GPU cloud platform for AI solution
• Strong problem-solving skills and the ability to handle complex technical challenges.
• Exposure to software development would be an added advantage.
• Deep understanding of the architectural principles for cloud-based platforms that include SaaS, PaaS, multi-tenancy, infrastructure as code, and continuous availability with hand-on.
• Excellent verbal, written, and presentation skills in English.
• Experience in stakeholder management and client engagement, ability to collaborate effectively across various cross-functional teams and groups.

Desirable qualifications

• Understanding how AI and HPC workloads interact with both GPU HW and SW infrastructure.
• Knowledge with orchestration platforms like Kubernetes, SLURM using ML frameworks.
• System-level experience specifically GPU-based systems.
• Understanding of how collective communications (MPI, RDMA, and NCCL) works on GPU cluster, as well as an understanding of GPU specific aceleration works.

Understanding of AI & HPC networking technologies such as InfiniBand, RoCE, DPUs.