Co-design the hardware-software interface for AMD Instinct GPUs across ROCm, IREE, and the Linux kernel DRM/GEM layer.
Originally posted November 16, 2026 by the employer.
About the role
The AMD AI Group is seeking a Senior Silicon Design Engineer to co-design the hardware-software interface for AMD Instinct GPUs. You will model and evaluate interactions across GPU microarchitecture, AI frameworks, and the ROCm software stack to enhance performance and energy efficiency.
What you'll do
- Build analytical and simulation-based performance and power models connecting GPU microarchitecture decisions to AI training and inference workload behavior.
- Evaluate proposed hardware features, cache coherence protocols, and memory subsystem changes by providing software-grounded feedback to AMD GPU hardware and firmware teams.
- Write and optimize GPU kernels, compute runtime paths, and compiler passes targeting AMD GPU microarchitecture.
- Contribute to the IREE compiler and the ROCm stack, addressing backend optimizations, runtime API extensions, and firmware command packet design.
- Work at the Linux kernel DRM/GEM layer on GPU compute infrastructure to improve asynchronous execution and lower latency.
- Develop reproducible benchmarks and regression infrastructure to validate speedups and workload characterization across compiler, runtime, and driver updates.
What you'll need
- Experience working across the compute stack spanning the Linux kernel DRM/GEM layer, the ROCm runtime, the IREE compiler, and AI frameworks such as PyTorch and Triton.
- Background in evaluating AI training and inference workloads against GPU microarchitectural targets.
- Ability to build performance and power models evaluating instruction execution, memory hierarchy behavior, networking, and compute-memory overlap.
- Experience optimizing GPU kernels and compute runtime paths for hardware utilization, pipelining, and latency reduction.
Nice to have
- Industry experience working at the boundary of GPU compute hardware and AI systems software.
- Strong computer architecture foundation including memory consistency and coherence models, GPU microarchitecture, and pipelining.
- Proficiency in C and C++ in system-level contexts, and Python for tooling, analysis, and framework-level work.
- Experience with GPU programming via CUDA and/or HIP, including authoring compute kernels beyond standard library calls.
- Familiarity with Linux kernel internals relevant to GPU compute, including the DRM/GEM subsystem, kernel-mode driver interfaces, and memory management.
- Direct experience with AMD Instinct GPU architectures (MI300X, MI350X, MI355X) and the ROCm stack.
- Working knowledge of compiler infrastructure including LLVM/MLIR-based flows, backend code generation, or IREE.
- Familiarity with AI model architecture concepts such as quantization techniques, KV cache, attention mechanisms, mixture-of-experts routing, or state space models.
- Bachelor's or Master's degree in Computer Science, Computer Engineering, or a related field.
Skills: ROCm, GPU microarchitecture, IREE, PyTorch, Linux kernel, performance modeling
More semiconductor jobs: AMD jobs
First seen in our index 0 days ago, well below the 47-day median for Infrastructure/Platform Software roles.
Infrastructure/Platform Software · GPU Software Stack
|
Open roles in category
338
|
Median days since first seen
47 d
|
Category median salary (all employers)
$220k
|
See the full market breakdown ▾Category comparison, skills in demand, and who else is hiring
| Metric | AMD | All employers we track in this specialty (338 roles · 24 employers) |
|---|---|---|
| Open roles in this specialty | 74 | 338 |
| Open roles in the wider Software, Firmware & Systems family | 381 | 6787 · 180 employers |
| Median days since first seen | 22 d | 47 d (+25 d vs this employer) |
| Median salary (USD postings) | n/a | $220k |
Skills observed across this category: ROCm, GPU microarchitecture, IREE, PyTorch, Linux kernel, performance modeling
Who's hiring in this category
- NVIDIA · 167 open roles · median 75 d
- AMD (this employer) · 74 open roles · median 26 d
- Intel Corporation · 29 open roles · median 27 d
- Qualcomm · 25 open roles · median 68 d
- Arm Holdings · 10 open roles · median 19 d
- Ampere Computing · 5 open roles · median 14 d
How we counted: 338 open Infrastructure/Platform Software (GPU Software Stack) roles from 24 employers tracked in the SemiconductorJobs index, counted 16 Nov 2026. Specialty figures count only roles carrying this exact specialty label, so an employer's related work in neighbouring specialties is not included there, and is counted in the wider Software, Firmware & Systems family row. Days are counted from when our index first saw each posting, not from the employer's original posting date.