Training Infrustructure Engineer

Pivot Search; Singapore; Vor Ort; SGD 180.000 - 300.000; Vollzeit

Zusammenfassung

Pivot Search is seeking a dedicated ML Systems Engineer to build the training systems behind our Large Physics foundation Model. You will partner with researchers to move experiments from prototype to scalable production.

Responsibilities include designing and optimizing distributed training across thousands of GPUs, researching parallelisation and precision trade-offs, and building robust checkpointing and reproducibility infrastructure to support rapid research iteration.

Qualifikationen

• Must have demonstrated ownership of large-model distributed training.
• Deep understanding of parallelism, memory optimization and mixed precision.
• Ability to investigate performance from framework internals to kernels and collectives.

Aufgaben

• Design, implement and optimise distributed training across thousands of GPUs.
• Research parallelisation and numerical precision trade-offs for novel architectures.
• Profile and debug low-level GPU operations to improve throughput and utilisation.
• Build checkpointing, fault-tolerance and reproducibility systems for rapid research iteration.
• Collaborate with researchers to scale new model architectures.

Kenntnisse

Distributed training

Parallelism & memory

Performance profiling

ML frameworks familiarity

Jobbeschreibung

THE ROLE Build the training systems behind our Large Physics foundation Model. Physical data and novel architectures demand new approaches beyond the standard language-and-vision training stack. You will partner with researchers to bring experiments from prototype to scale.

Responsibilities

• Design, implement and optimise distributed training across thousands of GPUs.
• Research parallelisation and numerical precision trade-offs for novel architectures.
• Profile and debug low-level GPU operations to improve throughput and utilisation.
• Build checkpointing, fault-tolerance and reproducibility systems that hold up through rapid research iteration.
• Collaborate with researchers to scale new model architectures.

Required Skills

• Demonstrated ownership of large-model distributed training with frameworks such as FSDP, DeepSpeed, Megatron, PyTorch or JAX/XLA.
• Deep understanding of parallelism, memory optimisation, mixed precision and communication overlap.
• Ability to investigate performance from framework internals down to kernels and collectives.
• Resourcefulness, fast execution and confidence learning unfamiliar domains.

Preferred Skills

• Open-source contributions to ML infrastructure, such as PyTorch, Megatron-LM, DeepSpeed or XLA.

*This role supports relocation to Singapore.*