AI Hardware Simulation Engineer

sygaldry-technologies; San Francisco (CA); On-site; USD 140,000 - 210,000; Full time

Benefits offered by this job

Visa Sponsorship

Compensation and equity

Health coverage

Offsite socials

Unlimited PTO

Job summary

Sygaldry Technologies is seeking a candidate to model and simulate high-performing AI hardware. The role interfaces with multiple internal teams and external partners to advance a hybrid architecture.

You will work with estimation tools and GPUs, CPUs, and novel accelerators in a distributed setup. Strong background in modeling or simulation within AI workloads, plus experience with C++ or SystemC, will help you contribute to a rigorous, scalable platform.

Qualifications

• Contributed to the modeling and simulation environment of a distributed AI acceleration architecture.
• Modeled complex training, inference and reasoning workflows in a heterogeneous AI cluster of GPUs, CPUs and novel accelerators.
• Exposure to the simulation environments used by the GPU/XPU supplier, hyperscalers and/or frontier labs for simulation AI workloads on various system and hardware architectures.
• Familiarity with architecture modeling and distributed simulation platforms, such as Gem5, ASTRA-sim and such.
• Experience with C++ or SystemC behavioral models and timing-aware behavioral models.

Responsibilities

• Contribute to and improve the ASTRA-sim simulation environment for our hypbrid architecture
• Interface with a variety of internal teams and external collaboration partners
• Work with internal teams that have developed estimation tools to help them integrate these tools into the simulation & modeling framework

Job description

Sygaldry Technologies is seeking a candidate to model and simulate high-performing AI hardware. The role interfaces with multiple internal teams and external partners to advance a hybrid architecture.

You will work with estimation tools and GPUs, CPUs, and novel accelerators in a distributed setup. Strong background in modeling or simulation within AI workloads, plus experience with C++ or SystemC, will help you contribute to a rigorous, scalable platform.