Mojo Hardware Enablement Engineer

Modular, a Qualcomm company; United States; Hybrid; USD 148,000 - 270,000; Full time

Benefits offered by this job

Healthcare coverage

Retirement plans

RSU grants

Employee stock purchase plan

Paid time off

Wellbeing resources

Learning & development

Team on-sites

Job summary

Modular, a Qualcomm company, is building the next generation AI platform. We’re expanding the Hardware Enablement team to support new accelerator platforms across the software stack, from Mojo kernels to the graph compiler and MAX model serving.

The role involves implementing architecture support, optimizing kernels, and collaborating with hardware partners to improve portability and tooling. Hybrid work with relocation assistance is available for those in the US and Canada.

Qualifications

• 2+ years of experience in high-performance computing or compiler engineering.
• Proficiency in C++ and experience with large multi-component software systems.
• Hands-on experience with at least one heterogeneous programming model (CUDA, SYCL, OpenCL).
• Some exposure to non-GPU accelerator architectures is a strong plus.

Responsibilities

• Implement and validate support for new hardware architectures across the Modular stack.
• Write and optimize Mojo kernels for accelerator architectures, focusing on correctness and performance.
• Contribute to portability infrastructure, tooling, and debugging workflows for new target hardware.
• Collaborate with hardware vendor engineers to build integration tests and triage issues.
• Develop knowledge of hardware platforms and share findings via demos and write-ups.
• Participate in on-site/company events to foster collaboration.

Skills

C++

HPC

Compiler engineering

CUDA

Tools

CUDA

SYCL

OpenCL

Job description

Modular, a Qualcomm company, is building the next generation AI platform. We’re expanding the Hardware Enablement team to support new accelerator platforms across the software stack, from Mojo kernels to the graph compiler and MAX model serving.

The role involves implementing architecture support, optimizing kernels, and collaborating with hardware partners to improve portability and tooling. Hybrid work with relocation assistance is available for those in the US and Canada.