below marketmost similar roles pay hereabove market
This role pays more than
57%
of similar roles. Most pay
$190,000–$235,750
— the blue band above.
At the midpoint, this role pays about
$218k
versus about
$213k
for comparable roles.
Based on 240 similar postings.
Employer
About Amd
AMD (Advanced Micro Devices) is a semiconductor company that develops high-performance processors, graphics cards, and adaptive computing solutions for gaming, data centers, and embedded markets. Industry: Semiconductors
Amd currently has
496 open roles
on FindRole.
Listed pay typically runs
$172,000–$258,000
across 496 roles with salary data.
The Senior GPU Performance Software Engineer joins the Datacenter Performance Group to drive performance analysis, SW/HW co-design, and architecture evaluation for next-generation GPU-based AI systems. This role involves analyzing AI workloads to identify bottlenecks, evaluating new architectural concepts, and prototyping emerging hardware features. The engineer will develop performance studies, benchmarks, and software prototypes to assess architectural tradeoffs across single-GPU, multi-GPU, and rack-scale environments. Key responsibilities include investigating system-level performance regarding communication, synchronization, and memory. Required skills include C/C++ software development, computer architecture knowledge, and experience with GPU software stacks, runtimes, drivers, or firmware. Candidates should be proficient with profiling, debugging, and tracing tools, and have experience with distributed systems using frameworks like RCCL or MPI to solve complex AI inference performance problems.
Analyze AI workloads on current and future GPU systems to identify performance bottlenecks and optimization opportunities.
Evaluate and prototype next-generation GPU features through data-driven performance analysis and SW/HW co-design.
Develop performance studies, benchmarks, and software prototypes to assess architectural tradeoffs and hardware impacts.
Investigate system-level performance across single-GPU, multi-GPU, and rack-scale environments.
Conduct pre-silicon architecture evaluation and post-silicon performance validation to drive hardware and software improvements.
Analyze communication, synchronization, memory, and workload distribution effects in large-scale systems.
What we're looking for
Bachelor's, Master's, or PhD in Computer Science, Computer Engineering, Electrical Engineering, or equivalent.
Strong software development experience in C/C++ with a focus on performance-sensitive systems (preferred).
Strong understanding of computer architecture, performance analysis, and optimization techniques (preferred).
Experience with GPU architectures and GPU software stacks, including programming models, runtimes, drivers, firmware, or performance tools (preferred).
Experience with distributed and multi-GPU systems, including communication and synchronization technologies such as RCCL or MPI (preferred).
Experience using profiling, debugging, and tracing tools to analyze complex software and system performance issues (preferred).
Demonstrated ability to leverage AI-assisted development tools, intelligent agents, and automation techniques to solve complex technical problems (preferred).
Familiarity with AI/ML workloads, GPU kernel development, or AI inference frameworks (preferred).
CUDA
C++
C
GPU Architecture
Parallel Programming
High-Performance Computing
TensorRT
TensorRT-LLM
cuTile
AI
Linear Algebra
Monte Carlo simulations
Inference Optimization
Profiling
Algorithms
Software Design