below marketmost similar roles pay hereabove market
This role pays more than
92%
of similar roles. Most pay
$179,575–$235,750
— the blue band above.
At the midpoint, this role pays about
$255k
versus about
$208k
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.
AI Engineer, Recursive Self-Improvement for Compute joins the team to build recursive self-improvement systems for compute. This role focuses on creating agentic and learning-driven optimization loops where AI proposes, verifies, and learns from improvements to compute workloads and hardware engineering workflows. You will develop systems that generate, compile, test, benchmark, and profile candidate improvements with minimal human intervention. Key responsibilities include designing feedback systems, reward modeling, and building reusable tools for hardware-aware optimization. The role requires proficiency in Python and systems languages like C++, C, HIP, or CUDA, alongside experience with PyTorch, JAX, and reinforcement learning. You will solve complex engineering problems involving GPU and CPU performance, compilers, and automated program repair to improve latency, throughput, and power efficiency across various compute domains.
What does a AI Engineer earn in California?
Median $246150 from 100 postings across 17 companies.
Build agentic and learning-driven optimization loops for compute workloads and hardware engineering workflows.
Develop systems to generate, compile, test, benchmark, and profile candidate improvements with minimal human intervention.
Convert high-value engineering workflows into verifiable tasks with clear graders, harnesses, and metrics.
Collaborate with AI researchers on reward design, reward shaping, and long-horizon optimization loops.
Design feedback systems that accumulate data from successful attempts, failed attempts, profiler traces, and validation logs.
Improve iteration speed through staged validation, caching, parallel execution, and faster feedback paths.
Build reusable tools and patterns that generalize across multiple compute and hardware optimization domains.
Mentor other engineers and set the technical direction for self-improving AI systems for compute.
What we're looking for
Bachelor's degree in Computer Science, Computer Engineering, Electrical Engineering, Machine Learning, or related field, or equivalent practical experience.
Master's degree preferred; PhD is a plus, especially in AI systems, reinforcement learning, compilers, GPU computing, or hardware/software co-design.
Strong software engineering experience in Python and at least one systems language such as C++, C, HIP, or CUDA (preferred).
Experience building AI, ML, agentic, optimization, or automation systems evaluated with objective metrics (preferred).
Experience with GPU kernels, ROCm/HIP, CUDA, Triton, PyTorch, JAX, TensorFlow, or distributed training/inference systems (preferred).
Experience with reinforcement learning, post-training, reward modeling, automated program optimization, or agentic coding systems (preferred).
Familiarity with CPU performance engineering, compiler optimization, benchmarking, profiling, or math libraries (preferred).
Exposure to hardware design, simulation, formal verification, performance/power/area analysis, or hardware/software co-design (preferred).
Python
PyTorch
JAX
TensorFlow
Reinforcement Learning
LLM Agents
Kubernetes
Ray
Slurm
CUDA
ROCm
HIP
Distributed Systems
Data Pipelines
Model Serving
Profiling
Compilers
Simulators
CUDA
C++
Python
Deep Learning
Distributed Computing
PyTorch
JAX
TensorRT
vLLM
Triton
XLA
NCCL
MPI
UCX
Kernel Optimization
Model Optimization
FP8
INT8
Generative AI