Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
развернуть
$160к-220к/год
$200к$230кмедиана$260к
По похожим ролям (50) ~ $230к/год
На СНГ рынке: $3.8к/мес ($2.4к - $5.4к)
88
Отличная вакансия
развернуть
Роль дает четкую ответственность за сложную систему симуляции для ИИ, а продуктовые планы и явно описанные льготы выглядят сильными; требование иметь разрешение на работу в США ограничивает доступ.
Только США
Актуальный стек
Инфраструктура ИИ
Зона ответственности
Сильный пакет льгот
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Distributed Systems Engineer, Multi-Chip Simulation (AI): Building and scaling multi-chip simulators for AI inference systems, including host-to-chip and chip-to-chip communication, with an accent on distributed protocols, timing, failure modes, and simulation fidelity. Focus on composing functional, cycle-level, and hardware-emulation models, scaling simulations to full racks, and correlating results with emulation and hardware measurements.
Location: Remote (United States)
Base salary: Mid: $160,000–$220,000; Senior: $220,000–$270,000; Principal: $270,000–$350,000 per year.
Company
hirify.global develops AI inference accelerators and multi-chip systems designed for low-latency, high-throughput large language model inference.
What you will do
Design and build host-to-chip and chip-to-chip simulators covering drivers, links, protocols, transfers, and communication fabrics.
Model transfer timing, collectives, ordering, back-pressure, congestion, and partial failure against real hardware.
Own the architecture of the multi-chip simulation stack as chips, interconnects, and deployment topologies evolve.
Compose functional, cycle-level, and hardware-emulation simulators into a coherent multi-chip simulation.
Scale simulations from a few chips to full racks for daily CI use.
Correlate simulated behavior with emulation and hardware measurements while maintaining regression coverage and documentation.
Requirements
5+ years of experience as a distributed systems engineer building and operating software coordinating many machines or accelerators through RPC, message passing, consensus, or collective communication.
Strong understanding of distributed-software failure modes, including races, partial failure, ordering and consistency, timeouts and retries, back-pressure, and deadlock.
Experience debugging timing-dependent bugs across hosts, NICs, or accelerators using traces and logs.
Sound judgment in selecting simulation fidelity and balancing accuracy against runtime.
Strong Rust skills, or expert-level C/C++ skills with willingness to work in Rust daily, plus proficiency in Python for tooling and analysis.
Excellent communication skills and the ability to explain simulation results, limitations, and recommended software or hardware changes.
Must be currently authorized to work in the United States.
Nice to have
Experience building simulators, emulators, or test harnesses for distributed or networked systems.
Hands-on experience with multi-GPU or multi-accelerator inference or training, including NCCL-style collectives, tensor, pipeline, and expert parallelism, and host-device runtimes.
Familiarity with Cadence Palladium, Synopsys ZeBu, or FPGA prototyping.
Culture & Benefits
Ownership of the simulation-stack architecture supporting distributed software across racks of chips before hardware exists.
Cross-functional work with chip modeling, SDK, and runtime teams.
Remote-first work with a company-provided computer and home-office setup.
Company-paid medical, dental, vision, life, and disability coverage, with additional supplemental coverage options.
Unlimited paid time off, 13 paid company holidays, equity, and a 401(k) with company matching.
No new visa sponsorship is available; H-1B transfers may be facilitated for eligible candidates.
Будьте осторожны
Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →