This role pays more than
58%
of similar roles. Most pay
$162,000–$235,750
— the shaded band above.
At the midpoint, this role pays about
$197k
versus about
$199k
for comparable roles.
Based on 239 similar postings.
Employer
About Nvidia
Nvidia is a leading designer of graphics processing units (GPUs) and system-on-chip units, powering gaming, professional visualization, data centers, and artificial intelligence workloads. Industry: Semiconductors & AI Computing
Nvidia currently has
896 open roles
on FindRole.
Listed pay typically runs
$184,000–$287,500
across 876 roles with salary data.
As a Solutions Architect, Inference Deployments, you will join a team focused on rolling out and enhancing AI inference solutions at scale using GPU technology and Kubernetes. You will collaborate with engineering, DevOps teams, and customers to build inference pipelines, distribute tasks among GPU workers, and orchestrate disaggregated inference for complex workloads. Your daily work involves accelerating inference pipelines using TensorRT-LLM, vLLM, SGLang, and NVIDIA Dynamo while providing technical leadership and mentorship to resolve complex deployment issues. You will utilize tools like Triton Inference Server, NVIDIA GPU Operator, NIM Operator, and Multi-Instance GPU partitioning. The role requires expertise in solving sophisticated GPU allocation, memory hierarchies, and low-latency networking using RDMA and UCX. You will focus on the technical challenge of tuning large language models for low-latency inference within enterprise environments to ensure seamless production integration.
What does a Solutions Architect earn in California?
Median $235750 from 45 postings across 8 companies.
Build inference pipelines using tools like NVIDIA Dynamo to distribute tasks among GPU workers for improved efficiency.
Orchestrate disaggregated inference workloads using Kubernetes in collaboration with DevOps teams.
Accelerate inference pipelines using TensorRT-LLM, vLLM, SGLang, and other backends.
Provide technical leadership and mentorship to customers during the deployment of disaggregated inference systems.
Resolve complex technical issues related to GPU allocation, memory hierarchies, and low-latency networking.
Tune large language models for low-latency inference in enterprise environments.
Implement model optimization techniques such as quantization and speculative decoding.
What we're looking for
At least 5 years of experience in Solutions Architecture with a track record of deploying distributed systems and AI inference workloads on Kubernetes.
Experience with NVIDIA Dynamo, Triton Inference Server, or TensorRT-LLM for model optimization and serving.
Expertise in GPU orchestration using NVIDIA GPU Operator, NIM Operator, and Multi-Instance GPU (MIG) partitioning.
Proficiency in solving complex GPU allocation, memory hierarchies, and low-latency networking like RDMA and UCX.
Demonstrated success tuning large language models for low-latency inference in enterprise environments.
A Bachelor of Science in Computer Science, Engineering, or equivalent experience.
Knowledge of transformer neural networks and acceleration technologies such as quantization and speculative decoding.
NVIDIA Certified AI Engineer or similar professional credentials.