This is a fully remote position, open to applicants in California, +1 more state.

📋 Description

• Collaborate in designing innovative architectures for NVIDIA's data processing platforms.

• Interact with end users, open-source communities, ISVs, storage partners, and CSPs to integrate customer feedback into roadmaps and reference architectures.

• Lead comprehensive performance optimization efforts from CUDA kernels and accelerated libraries to system-level IO, shuffle, compression, and data transfer.

• Partner with CPU, GPU, and platform architects on hardware-software co-design initiatives.

• Create and implement constraint-based frameworks and optimization solvers to address design trade-offs.

• Establish benchmarking frameworks to assess performance sensitivities and direct hardware-software co-optimization.

• Examine production workloads to uncover compute motifs, IO patterns, and optimization prospects throughout the platform stack.

⛳️ Requirements

• MS or PhD in Computer Science, Computer Engineering, Electrical Engineering, or a related discipline, or equivalent experience.

• Over 8 years of experience in distributed data platforms, databases, systems engineering, performance architecture, accelerated computing, or HW/SW co-design.

• In-depth knowledge in one or more areas including data processing systems, accelerated computing, high-performance IO, CPU/GPU architecture, or large-scale performance analysis.

• Experience in evaluating system-level trade-offs across compute, IO provisioning, memory subsystem development, network topology, and storage tiering is highly advantageous.

• Capability to convert complex system analyses into clear and actionable recommendations.

• Proficiency in collaborating across matrixed organizations.

• Practical experience with GPU-accelerated data processing is a plus.

• Strong familiarity with structured data engines and/or vector database systems is a plus.

• Experience with mathematical optimization solvers and optimization principles is advantageous.

• Background in network/storage architecture is an added benefit.