AI Inference Performance Engineer for NVIDIA GPUs

YOH Services LLC; Santa Clara (CA); On-site; USD 250,000 - 300,000; Full time

Benefits offered by this job

Medical benefits

Health Savings Account (HSA)

401K Retirement Savings Plan

Life & Disability Insurance

Referral Bonus Programs

Job summary

YOH Services LLC is seeking a hands-on Performance/Benchmark Engineer to characterize and optimize AI workloads on large-scale NVIDIA GPU infrastructure. The role emphasizes GPU compute performance, AI inference, and system-level benchmarking within an architecture team.

Responsibilities include developing benchmarks, evaluating models for bottlenecks, and collaborating with cross-functional teams to improve end-to-end cluster performance. Located in California with on-site duties.

Qualifications

• Hands-on experience with NVIDIA GPU compute platforms and AI/ML benchmarking.
• Understanding of AI inference, performance modeling, and GPU architecture.
• Experience with DGX, B200/B300, H100/H200, Blackwell, Hopper or comparable GPUs.
• Ability to analyze latency, throughput, memory bandwidth, and multi-GPU scaling.
• Strong Python scripting and automation skills.

Responsibilities

• Develop and run performance benchmarks for AI inference across GPU systems.
• Characterize performance on DGX and related platforms, analyzing throughput and latency.
• Identify bottlenecks and propose system or architecture improvements.
• Build benchmarking methodologies, automation, and reporting frameworks for repeatable results.
• Collaborate with architecture, compute, networking, and software teams to optimize end-to-end performance.

Skills

NVIDIA GPU compute

AI inference benchmarking

GPU architecture knowledge

Python scripting

Tools

CUDA

NCCL

TensorRT

Nsight

Job description

YOH Services LLC is seeking a hands-on Performance/Benchmark Engineer to characterize and optimize AI workloads on large-scale NVIDIA GPU infrastructure. The role emphasizes GPU compute performance, AI inference, and system-level benchmarking within an architecture team.

Responsibilities include developing benchmarks, evaluating models for bottlenecks, and collaborating with cross-functional teams to improve end-to-end cluster performance. Located in California with on-site duties.