AI Infrastructure Performance Engineer | HPC & CUDA

Acceler8 Talent; Massachusetts; Hybrid; USD 150,000 - 210,000; Full time

Benefits offered by this job

Competitive salary

Benefits

Meaningful equity

Job summary

Acceler8 Talent is seeking a Member of Technical Staff in Cambridge, MA, hybrid 3 days per week, with remote options across the USA & Canada. You will join a team building high-performance AI infrastructure, focusing on identifying bottlenecks, implementing fixes, and proving improvements in distributed training, large-scale inference, and HPC environments.

The role emphasizes profiling, end-to-end optimization, and delivering tangible performance gains, with competitive salary, benefits, and

Qualifications

• 3+ years of whole-stack performance work.
• 3+ years of systems programming in C++, CUDA or Triton.
• Experience profiling complete workloads and fixing bottlenecks.
• Experience improving training/serving/simulation workloads across nodes.

Responsibilities

• Analyze performance bottlenecks across distributed workloads.
• Implement optimizations to improve throughput and latency.
• Collaborate with AI infra teams on scalable performance improvements.

Skills

Whole-stack performance

C++/CUDA/Triton

Profiling workloads

Cross-node optimization

Job description

Acceler8 Talent is seeking a Member of Technical Staff in Cambridge, MA, hybrid 3 days per week, with remote options across the USA & Canada. You will join a team building high-performance AI infrastructure, focusing on identifying bottlenecks, implementing fixes, and proving improvements in distributed training, large-scale inference, and HPC environments.

The role emphasizes profiling, end-to-end optimization, and delivering tangible performance gains, with competitive salary, benefits, and