Junior GPU Software Engineer

About the Role:

You will join OXMIQ as a Junior GPU Software Engineer (1–4 years experience) in Hyderabad / Secunderabad, Telangana. You will work on an early-career team to build, test, and optimize kernels, runtime, and enablement layers for high-performance AI workloads.

• Develop, test, and optimize GPU software within OxPython across kernel, runtime, and framework integration layers.
• Write and tune GPU kernels (CUDA, Triton-based) for operations like GEMM, attention, convolution, and memory-bound tasks.
• Profile workloads to identify bottlenecks and improve hardware utilization.
• Apply SIMT concepts including warp scheduling, divergence control, occupancy optimization, and memory hierarchy management.
• Enable new models on Oxmiq GPU IP using pre-silicon and emulation platforms.
• Collaborate with compiler engineers (MLIR/LLVM) to map kernels onto hardware.

• B.E./B.Tech/M.S. in Computer Science, Engineering, or related fields.
• Strong C/C++ and systems programming fundamentals.
• Hands-on GPU programming experience with CUDA (via coursework, research, or projects).
• Understanding of parallel computing: threads/warps, synchronization, memory hierarchies.