Remote Senior Compiler Engineer for Custom AI Accelerators

5V Tech; United States; On-site; USD 180,000 - 270,000; Full time

Benefits offered by this job

Significant equity

Bonus

Comprehensive health insurance

Job summary

5V Tech, a US-based AI computing company, seeks a Senior Compiler Engineer to shape the compiler backends for custom accelerator silicon. This hands-on role offers high ownership and direct influence on architecture and performance across the toolchain.

You will drive design and implementation of backends, code generation, and runtime integration, with a strong emphasis on GPU/accelerator kernels, LLVM/MLIR, and low-level optimisations.

Qualifications

• 8+ years in compiler engineering with strong backends experience.
• Deep hardware knowledge alongside compiler background.
• Experience writing or optimising GPU/accelerator kernels.
• Working knowledge of runtimes and the ML software stack.
• Strong C++ and Python with solid computer architecture fundamentals.
• Exposure to kernel level DSLs such as Triton, CUTLASS, or TVM.

Responsibilities

• Design and implementation of compiler backends targeting custom accelerator hardware.
• Code generation, instruction selection, scheduling, and register allocation.
• Runtime and ML software stack work that sits alongside the compiler.
• Low level performance optimisation across execution, memory movement, and parallelism.
• Benchmarking and profiling to close gaps against hand tuned baselines.
• Automated and tooling driven approaches to compiler and performance optimisation.

Skills

C++

Python

High-performance computing

GPU/accelerator awareness

CPU architecture fundamentals

Education

MS or PhD in CS/EE

Tools

LLVM

MLIR

Triton

CUTLASS

TVM

PyTorch internals

Job description

5V Tech, a US-based AI computing company, seeks a Senior Compiler Engineer to shape the compiler backends for custom accelerator silicon. This hands-on role offers high ownership and direct influence on architecture and performance across the toolchain.

You will drive design and implementation of backends, code generation, and runtime integration, with a strong emphasis on GPU/accelerator kernels, LLVM/MLIR, and low-level optimisations.