Systems Performance Engineer, AI Solutions Lab
About The Position
The Systems Performance Engineer will work on Marvell’s AI Solutions Lab within the Solutions Architecture team in the Platform organization. The team validates AI and cloud platforms across compute, networking, and memory fabrics, demonstrating their performance at rack scale under customer-relevant conditions. Its work includes running AI benchmarking experiments, operating GPU/HPC lab clusters, validating Teralynx switching and AI fabrics, and building live demonstrations for trade shows. The role focuses on end-to-end performance optimization across AI serving and training stacks. The engineer will identify bottlenecks across application flows, compute, memory, communications, GPU kernels, network transport, and data/storage pipelines; develop reproducible benchmark methodologies; analyze serving-stack internals; and turn results into recommendations and performance narratives. The role works with NRE/SRE and network or fabric architects to address competitive gaps, and prepares benchmark artifacts and presentations for customers, executives, and demonstrations. The team’s platform scope spans networking, integrated optical platforms such as CPO and NPO, and memory-fabric products such as CXL and photonic fabric memory appliances. The position requires a bachelor’s or master’s degree in Computer Science, Computer Engineering, or a related field; a PhD is a plus. The expected base pay range is $186,900–$275,000 per annum. The role is eligible for Marvell bonus and equity plans, including a new-hire equity grant and annual equity refreshers. Marvell also describes benefits across financial well-being, family support, mental and physical health, and recognition.
Requirements
- Experience with AI serving frameworks such as vLLM, SGLang, TensorRT-LLM, or similar platforms.
- Strong background in RCCL/NCCL, RDMA, communication-compute overlap, and performance profiling.
- Hands-on experience with GPU profiling tools such as Nsight Systems/Compute, PyTorch Profiler, or ROCm Compute Profiler to localize kernel-level bottlenecks.
- Demonstrated ability to diagnose end-to-end bottlenecks across application, framework, runtime, and infrastructure layers.
- Working knowledge of at least one distributed training framework, such as PyTorch DDP/FSDP, DeepSpeed, or Megatron-LM.
- Strong Python and Linux skills.
- Comfort working in containerized environments such as Docker to ensure benchmark reproducibility.
- Ability to translate technical analysis into clear recommendations and concise performance narratives.
- Bachelor’s or master’s degree in Computer Science, Computer Engineering, or a related field.
Nice To Haves
- Go experience.
- Kubernetes experience.
- Experience benchmarking distributed large-model inference and training systems.
- Familiarity with GPU kernel analysis, system tracing, and workload-specific optimization.
- Familiarity with serving-stack internals such as continuous batching, paged KV-cache attention, and quantization techniques such as GPTQ and SmoothQuant.
- Exposure to customer proofs of concept, competitive benchmarking, or externally consumed performance reporting.
- A PhD.
Responsibilities
- Drive end-to-end performance optimization across AI serving and training stacks using leading models and customer-relevant configurations.
- Identify bottlenecks across application request flow, compute, memory, communication layers, GPU kernels, network transport, and data/storage pipelines.
- Develop and execute reproducible, technically credible benchmark methodologies relevant to customer use cases.
- Profile serving-stack internals, including quantization, scheduling, KV-cache management, and disaggregated prefill/decode, to identify optimization targets.
- Analyze results and develop performance narratives to guide kernel, runtime, communication, and system-level optimizations.
- Work with NRE/SRE and network or fabric architects to close competitive gaps through targeted tuning.
- Create performance summaries, visuals, and benchmark artifacts for customer presentations, executive reviews, and trade-show demonstrations.
Benefits
- Bonus and equity plans, including a new-hire equity grant and annual equity refreshers.
- Employee stock purchase plan with a 2-year look back.
- Family support programs.
- Mental health resources.
- Recognition and service awards.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
