Full-Chip Performance Modeling Architect (AI/LLM)

Bitdeer Group; Singapore; On-site; SGD 80,000 - 120,000; Full time

Job summary

Bitdeer Group in Singapore is looking for a performance modeling expert to help analyze and optimize chip performance through modeling techniques. You will be responsible for building and maintaining cycle-approximate models that guide architecture decisions concerning tile count, NoC bandwidth, and local memory.

Your expertise in C++ and Python, alongside strong collaboration skills, will be essential. The role offers an opportunity to work closely with various teams and make impactful decisions in the performance of advanced chips.

Qualifications

• 5+ years in performance modeling or architecture for SoCs, accelerators, GPUs, or CPUs.
• Strong skills in C++ and Python; experience with SystemC or comparable frameworks is needed.
• Demonstrated ability to model subsystems and translate results into architecture decisions.

Responsibilities

• Build and maintain performance models of the tile array, NoC, memory subsystem, and IO.
• Characterize LLM and CNN inference workloads; produce relevant analyses.
• Model NoC contention and memory bandwidth based on realistic traffic.

Skills

Performance modeling / architecture

C++

Python

SystemC

Job description

Bitdeer Group in Singapore is looking for a performance modeling expert to help analyze and optimize chip performance through modeling techniques. You will be responsible for building and maintaining cycle-approximate models that guide architecture decisions concerning tile count, NoC bandwidth, and local memory.

Your expertise in C++ and Python, alongside strong collaboration skills, will be essential. The role offers an opportunity to work closely with various teams and make impactful decisions in the performance of advanced chips.