Sr. Staff Distributed Systems Software Engineer, Collectives - TX, Austin, United States
Originally posted 28 October 2026 by the employer.
Optimize collective communication libraries and ROCm network software for multi-GPU systems.
About the role
As a member of the Collectives and Network Performance team, you will optimize collective communication and networking software for AMD GPUs. You will enhance communication libraries and training and inference performance across multi-GPU and multi-node systems, working directly with ROCm runtime and library teams.
What you'll do
- Design, implement, and optimize features in collective communication libraries across scale-up and scale-out networks. - Profile, tune, and analyze LLM, diffusion, multimodal, RecSys, and generative AI workloads on single-node and multi-node systems. - Drive features from design through delivery, including scoping, implementation, testing, and performance validation. - Develop and improve tools and infrastructure for benchmarking, performance estimation, regression tracking, and reporting. - Collaborate with internal GPU library, runtime, and ROCm teams, as well as open-source maintainers, to integrate code changes upstream. - Provide software performance data and analysis to inform hardware-software co-design discussions for future scale-up networks, NICs, and scale-out networks.
What you'll need
- Professional software development experience in Linux environments focusing on C++ development. - Knowledge of network, NIC, and GPU hardware architecture. - Hands-on experience profiling and optimizing distributed AI workloads, along with understanding modern training and inference optimization. - Master's degree in Computer Science, Computer Engineering, Electrical Engineering, or a related field (or a Bachelor's degree with equivalent experience).
Nice to have
- Strong knowledge of collective communication libraries such as RCCL, NCCL, or MPI, plus familiarity with NIC, RDMA, and scale-up and scale-out network architectures. - Experience in profiling, performance analysis, and bottleneck identification for distributed AI workloads. - Hands-on experience integrating optimized compute or communication into ML frameworks such as PyTorch, JAX, vLLM, or SGLang. - Strong proficiency in C++ and Python with experience in performance tuning, debugging, and test design. - Experience with large-scale, multi-node compute environments and optimizing AI workloads across clusters. - Senior level professional experience in technical software development focused on performance engineering, communication libraries, networking, or GPU/distributed systems. - PhD in Computer Science, Computer Engineering, Electrical Engineering, or a related field.
Skills: RCCL, NCCL, ROCm, collective communication, C++, MPI
This role has been open 0 days — well below the 55-day median for Infrastructure/Platform Software roles.
Infrastructure/Platform Software · GPU Software Stack
|
Open roles in category
258
|
Median days open
55 d
|
Median salary
$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 (258 roles · 19 employers) |
|---|---|---|
| Open roles in this specialty | 54 | 258 |
| Open roles in the wider Software, Firmware & Systems family | 387 | 6357 · 147 employers |
| Median days open | 57 d | 55 d (+2 d vs this employer) |
| Median salary (USD postings) | — | $220k |
Skills observed across this category: RCCL, NCCL, ROCm, collective communication, C++, MPI
Who's hiring in this category
- NVIDIA · 127 open roles · median 58 d
- AMD (this employer) · 55 open roles · median 57 d
- Intel Corporation · 29 open roles · median 15 d
- Qualcomm · 17 open roles · median 188 d
- Arm Holdings · 8 open roles · median 1 d
- Graphcore · 4 open roles · median 8 d
How we counted: 258 open Infrastructure/Platform Software (GPU Software Stack) roles from 19 employers tracked in the SemiconductorJobs index, counted 29 Oct 2026. Specialty figures count only roles carrying this exact specialty label, so an employer's related work in neighbouring specialties is not included there — it is counted in the wider Software, Firmware & Systems family row. Figures refresh nightly.