GPU Compiler Engineer - Cambridge (Perm)

Role and Responsibilities

Define GPU compiler software architecture and interfaces.

Development/implement GPU compiler pipeline, linking and various optimizations/transformations.

Collaborate with Driver team, HW team to implement new API & HW features.

Collaborate with Driver team, HW team to improve/tune performance & power consumption.

Execute & deliver to meet milestones/schedules.

Analyze and debug code generation issues.

Analyze and influence future GPU architectures.

Construct reliable & trustable relationships across teams internally & externally.

\n
• BS, MS or PhD degree or equivalent in Computer Science, Electrical or Computer Engineering or related majors.
\n
• Working experience on product compiler, compiler tools such as assembler, linker development.
\n
• Hands-on knowledge/experience on C/C++ programming.
\n
• Hands-on knowledge/experience on state-of-the-art compiler optimization/transformation techniques.
\n
• Hands-on knowledge/experience on LLVM.
\n
• Strong written & verbal communication skills.
\n
• Strong teamwork skill and execution capability.
\n
• Strong problem solving, root causing and debugging skills.
\n
• Strong self-driven skills.
\n
• Preferred skills:
\n
• Knowledge/experience on GPU/GPU pipeline and GPU architecture.
\n
• Knowledge/experience on SIMD architecture.
\n
• Knowledge/experience on D3D, Vulkan/OpenGL, OpenCL APIs.
\n
• Knowledge/experience on leveraging AI in compiler/tool/optimization development.
\n

JBRP1_UKTJ