The Role:We are building next-generation systems software for AMD GPUs with Rust as a
core technical direction. The work spans compilers, runtimes, low-level GPU
software, firmware, developer tooling, and methods for improving the safety and
correctness of complex hardware-software systems.
Rust is not incidental to this role. You will help establish how Rust is used
in performance-sensitive and high-trust parts of the GPU stack, including the
compiler and tooling support, system interfaces, engineering practices, and
validation methods required for production deployment. You will have the
opportunity to influence both near-term implementations and the longer-term
architecture for using Rust across current and future AMD platforms.
The Person:This role is for a deeply technical compiler and systems generalist who enjoys
working across abstraction layers. You may move between Rust source semantics,
rustc internals, intermediate representations, MLIR and LLVM, target-specific
code generation, object formats, runtimes, GPU ISA, firmware, and hardware
interfaces.
You should have strong judgment about correctness boundaries. Memory safety is
important, but production correctness also depends on concurrency, ordering,
ABI compatibility, device behavior, compiler transformations, and the
assumptions between software and hardware. You will help make those assumptions
explicit and build evidence that the resulting systems behave as intended.
The ideal candidate has deep expertise in at least one part of this stack and
the ability to become productive in unfamiliar areas. Experience with every
technology listed below is not expected. A research background is not
required, although the ability to read technical work, form testable
hypotheses, and turn promising ideas into reliable implementations is valuable.
The work sits at the intersection of Rust systems programming, production
compiler infrastructure, GPU architecture, firmware, program analysis, and
formal methods. Hardware-software co-design experience is highly valuable,
especially when compiler, runtime, firmware, or ISA choices must evolve
together.
Key Responsibilities:- Lead the design and implementation of production-quality Rust components for
low-level and performance-sensitive GPU systems. - Help make Rust a first-class engineering language across relevant parts of
the GPU software stack, including toolchain integration, libraries, coding
standards, diagnostics, testing, and deployment practices. - Design and implement compiler pipelines involving rustc, MIR, target-neutral
intermediate representations, MLIR dialects and transformations, LLVM IR,
and AMDGPU code generation. - Build analyses, transformations, legalization passes, and target-specific
lowering for GPU execution models, memory spaces, synchronization,
capabilities, and specialized hardware operations. - Define typed and auditable interfaces across compiler, host runtime, device
code, firmware, and hardware boundaries. - Work on ABI definition and validation, argument layout, calling conventions,
linker and loader behavior, ELF and code-object metadata, and source-to-binary
identity. - Develop or improve bare-metal and `no_std` Rust components where explicit
state, bounded behavior, fail-closed handling, and predictable hardware
effects are required. - Use testing, fuzzing, property-based testing, differential execution,
executable models, and hardware validation to find semantic gaps and prevent
regressions. - Apply program analysis or formal methods where they provide meaningful
assurance, including source-level invariants, state-machine properties,
refinement obligations, and clearly documented trusted boundaries. - Distinguish precisely among type-system guarantees, source-level proofs,
compiler evidence, binary validation, simulator results, and behavior
observed on hardware. - Partner with compiler, runtime, firmware, architecture, validation, security,
and performance teams to solve problems that cross organizational and
technical boundaries. - Write design documents, lead technical reviews, mentor engineers, and explain
complex language, compiler, and hardware behavior clearly. - Evaluate emerging compiler and verification technologies and integrate them
when they improve the safety, performance, maintainability, or development
velocity of production systems.
Preferred Experience:- Strong production Rust experience, including ownership and borrowing,
lifetimes, traits and generics, error handling, concurrency, FFI, and the
careful design of safe abstractions over unsafe operations. - Experience with low-level or constrained Rust, including `no_std`, embedded
systems, firmware, custom targets, allocators, volatile access, interrupt
handling, or architecture-specific code. - Strong C or C++ systems programming skills and the ability to work across
mixed-language codebases. - Experience with rustc internals, MIR, custom codegen backends, procedural
macros, compiler plugins, or other Rust toolchain infrastructure. - Experience with LLVM, MLIR, GCC, or another production compiler framework.
- Experience designing intermediate representations, dialects, analyses,
optimization passes, legalization, instruction selection, scheduling, or
target-specific code generation. - Understanding of compiler correctness issues such as undefined behavior,
aliasing, provenance, layout, calling conventions, memory models, and
source-to-target semantic preservation. - Ability to read and reason about low-level assembly, ISA encodings, compiler
output, and machine-level execution. - Experience with linkers, loaders, ELF or other object formats, code-object
metadata, DWARF, LLVM MC, disassembly, post-link analysis, or binary rewriting. - Understanding of GPU execution models, including waves or warps, SIMD/SIMT
execution, address spaces, registers, local or shared memory, caches,
synchronization, atomics, barriers, and asynchronous operations. - AMDGPU, GCN, RDNA, CDNA, ROCm, HIP, HSA, or AMDGPU LLVM backend experience.
- Experience with GPU runtimes, drivers, firmware, embedded processors, device
initialization, command processing, queueing, or hardware control paths. - Experience building secure or high-assurance systems that process untrusted
inputs or operate across privilege, memory-isolation, or hardware trust
boundaries. - Experience with program analysis, abstract interpretation, symbolic
execution, model checking, SMT solvers, or formal verification tools such as
Verus, Lean, Coq, Isabelle, Dafny, or TLA+. - Experience writing executable specifications, proving state-machine
invariants, or validating refinement between source, IR, binaries,
simulators, RTL, and hardware. - Strong validation instincts, including adversarial testing, differential
testing, property-based testing, fuzzing, reproducible evidence, and careful
treatment of assumptions and counterexamples. - Hardware-software co-design experience for ISA features, compiler-visible
architecture, memory systems, synchronization, security mechanisms, or
specialized compute pipelines. - Technical leadership experience: setting direction, resolving ambiguous
cross-layer problems, influencing partner teams, mentoring engineers, and
communicating tradeoffs clearly.
Preferred Academic Credentials:Bachelor's, Master's, or PhD in Computer Science, Computer Engineering,
Electrical Engineering, or a related field, or equivalent industry experience.
Research publications and advanced degrees are welcome but are not required.
LOCATION:San Jose, California
#LI-G11
#LI-HYBRID
Benefits offered are described: AMD benefits at a glance.
