Senior Applied Research Engineer, Accelerator Programming Model and Compiler at NVIDIA (Bengaluru)

Job Description

NVIDIA is hiring a Senior Applied Research Engineer for its accelerator programming model and compiler in Bengaluru to design the next-generation way of programming its Programmable Vision Accelerator (PVA) for both developers and AI coding agents, from high-level declarative abstractions down to an LLVM-based compiler backend.

At a glance

• Role: Senior Applied Research Engineer, Accelerator Programming Model and Compiler
• Location: Bengaluru
• Experience: 10+ years in low-level systems, accelerator, embedded or compiler/toolchain software
• Focus: PVA programming model, LLVM backend and agent-driven development
• Requisition: JR2022416
• Apply: open until filled (re-checked 12 October 2026)

What you will do

• Work on the next-generation PVA programming model to make optimised algorithms easier to build
• Enable AI agent-driven PVA development through declarative interfaces, faster compile and emulation, and better diagnostics
• Help define the architecture and features of the PVA SDK, runtime APIs and programming model
• Define benchmarks and evaluations for agent-generated PVA code and use them to improve compiler optimisation
• Develop the LLVM-based VPU compiler backend for the VLIW/SIMD architecture
• Research integration of PVA workloads into CUDA-based heterogeneous pipelines, and work with customers on real workloads

Several NVIDIA roles in Bengaluru centre on the Programmable Vision Accelerator (PVA), which the postings describe as a power-efficient, deterministic VLIW/SIMD processor used across NVIDIA’s DRIVE, Jetson and IGX platforms. Work on it spans algorithms, compilers, SDKs and future hardware.

This role also works on making the accelerator easy to program for AI coding agents as well as people: declarative interfaces, fast compile and emulation, and compiler feedback that helps agents produce optimised code.

What NVIDIA is looking for

• A BS, MS or PhD in computer science, computer or electrical engineering, or equivalent experience
• 10+ years building high-performance low-level systems, accelerator, embedded or compiler/toolchain software
• Experience building DSLs, compiler IRs or programming abstractions that hide hardware complexity without losing performance
• Compiler, debugger, linker or toolchain development, especially with LLVM
• Experience developing with coding agents (the posting names Claude Code, Cursor and OpenAI tools) and integrating developer tools with agent harnesses
• SIMD/VLIW programming, strong C++ with low-level debugging, and Linux or QNX

Nice to have

• CUDA, especially integrating accelerators into CUDA pipelines
• OpenCL, MLIR, Halide, TVM, Triton, graph compilers or image-processing DSLs
• ISO 26262 or IEC 61508
• Building agent harnesses and long-running multi-agent workflows

Pay and location

NVIDIA’s posting does not state a salary. The role is based in Bengaluru.

How to apply

Apply through the official NVIDIA job posting. NVIDIA’s posting does not give a closing date, so the role is open until filled; ResearchJobs.in will re-check this listing on 12 October 2026. Check the posting for location and work-authorisation details before you apply.

See all our artificial intelligence jobs, or browse more research jobs on ResearchJobs.in.