Performance Software Engineer — Kernels for AI Accelerators

arago; Paris; Sur place; EUR 90 000 - 120 000; Plein temps

Avantages offerts par ce poste

Stock options

Relocation bonus

Healthcare coverage

Pension contributions

25 days PTO

Résumé du poste

Arago is an AI and hardware company building the core computation software for custom AI accelerators. We connect host systems to onboard kernels and tooling for real-time debugging, driving chip bring-up and high-performance computation Kernels.

The Performance Software Engineer will develop low-level primitives in C/C++ and assembly, work across hardware interfaces, and help ensure deterministic, efficient software on constrained environments.

Qualifications

• Bachelor's/Master's/PhD in Computer Science, Computer Engineering, or a related field.
• Strong experience in C/C++ and assembly with understanding of mapping to machine instructions and hardware behavior.
• Bare-metal development experience including startup code, linker scripts, memory layout, MMIO, interrupts/exceptions, timers, synchronization, and peripheral access.
• Deep understanding of computer architecture (pipelines, caches, memory hierarchies, DMA, atomics).
• Hands-on experience with RISC-V and/or ARM architectures and ISA concepts.
• Experience debugging hardware/software interactions using GDB, JTAG, trace interfaces, simulators, emulators, or FPGA platforms.
• Ability to read hardware specs, RTL docs, and translate to robust software implementations.
• Strong performance-analysis and optimization skills including profiling and hand-tuned assembly where appropriate.
• Proficiency in English; French is a plus.

Responsabilités

• Develop and optimize low-level software primitives in C/C++, assembly, and architecture-specific intrinsics for RISC‑V and ARM.
• Implement performance-critical kernels, runtime components, and firmware for bare-metal environments.
• Collaborate with hardware teams to define and validate hardware–software interfaces, memory maps, and boot flows.
• Analyze and optimize software for latency, throughput, and memory footprint.
• Bring up and debug software on simulation, emulation, FPGA, and silicon platforms.
• Develop validation tests, benchmarks, and diagnostic tools for processor, DSP, accelerator, and SoC functionality.
• Contribute to architecture decisions around ISA extensions and memory hierarchy.

Connaissances

C/C++ programming

Assembly language

Bare-metal development

Computer architecture

RISC-V/ARM

Debugging hardware/software

GDB/JTAG/trace

English communication

Formation

Bachelor's degree in CS/EE

Outils

GDB

JTAG

Trace interfaces

Simulators

Emulators

FPGA platforms

Description du poste

Arago is an AI and hardware company building the core computation software for custom AI accelerators. We connect host systems to onboard kernels and tooling for real-time debugging, driving chip bring-up and high-performance computation Kernels.

The Performance Software Engineer will develop low-level primitives in C/C++ and assembly, work across hardware interfaces, and help ensure deterministic, efficient software on constrained environments.