Sr. Engineer, Kernel Development and Optimization

Job Title

Sr. Engineer, Kernel Development and Optimization

Role Summary

The Kernel Development and Optimization team develops performance-critical kernels to maximize throughput and efficiency of Tenstorrent AI hardware for ML and HPC workloads. This role implements and optimizes GPU-style kernels, contributes host-side orchestration, and collaborates with compiler, runtime, ML, and hardware teams to deliver production-quality performance. This is a hybrid role based out of Belgrade, Serbia.

Experience Level

Senior level. The company may assess candidates at different levels during interview and align offers accordingly; no specific years-of-experience were stated.

Responsibilities

Deliver and maintain high-performance kernels and supporting host code to achieve measurable throughput and correctness gains.

• Design, implement, and optimize GPU-style kernels (matrix multiply, attention primitives, data-movement operations).
• Own performance workstreams: identify bottlenecks, design experiments, and deliver measurable throughput improvements.
• Contribute host-side orchestration and parallelization strategies to integrate kernels into production runtimes.
• Develop micro-benchmarks, regression tests, and tooling to ensure correctness and sustained performance.
• Use profiling and benchmarking data to guide optimization decisions.
• Debug complex runtime- or kernel-level issues in large codebases.
• Collaborate closely with compiler, runtime, ML, and hardware teams during integration.

Requirements

Core technical requirements and preferred skills.

Must-have:

• Strong C++ systems engineering experience writing performance-critical or low-level software.
• Experience with concurrency, synchronization, and latency-hiding techniques.
• Proven ability to use profiling and benchmarking tools to drive optimizations.
• Experience debugging complex kernel or runtime-level issues in large codebases.
• Experience designing or optimizing kernels with compute vs. memory trade-offs.
• Effective collaborator across compiler, runtime, ML, and hardware teams.
• Ability to work hybrid based out of Belgrade, Serbia and comply with U.S. export-control eligibility requirements where applicable.

Nice-to-have:

• Experience with accelerator runtimes, low-level kernel toolchains, or AI/ML workloads.
• Familiarity with micro-benchmarking frameworks, regression testing, or AI-assisted kernel tooling.

Education Requirements

Not specified.

About the Company

Company: Tenstorrent

Headquarters: Austin, Texas, United States

Tenstorrent is a technology company focused on designing innovative computing solutions. They are known for their expertise in the development of advanced hardware, including ASICs and SoCs, aimed at enhancing performance and efficiency in various applications.