Datacenter & Agentic AI Workload Performance Optimization Engineer
$100k - $500k • Remote • Santa Clara, California, United States; Toronto, Ontario, Canada • Full-time
Posted 25d ago
About the job
Tenstorrent is seeking a Workload Performance Optimization Engineer to enhance the performance of software workloads on next-generation RISC-V platforms. This role involves optimizing modern datacenter and agentic AI workloads across various programming languages like Java, Python, and Rust. You will identify performance bottlenecks and develop optimizations at the software, compiler, runtime, and hardware levels to improve throughput, latency, and efficiency. The position bridges software runtimes, compilers, CPU microarchitecture, and RISC-V silicon, requiring profiling of real-world applications, investigating memory and concurrency, and leveraging RISC-V capabilities. You will also contribute to AI-assisted development and automated optimization workflows, directly impacting the RISC-V software ecosystem and future CPU architectures.
Responsibilities
- Optimize software workloads on next-generation RISC-V platforms.
- Identify performance bottlenecks in datacenter and agentic AI workloads.
- Develop software, compiler, runtime, and hardware-aware optimizations.
- Improve throughput, latency, and efficiency of AI workloads.
- Bring up and tune major software runtimes.
- Profile real-world applications and investigate memory/concurrency behavior.
- Explore optimizations using RISC-V Vector/Matrix capabilities and custom instructions.
- Work with AI-assisted development and automated optimization workflows.
- Influence the RISC-V software ecosystem and future CPU architecture.
Requirements
- Hands-on experience optimizing software on RISC-V or another modern CPU architecture.
- Strong understanding of the hardware/software boundary.
- Comfortable profiling complex systems, finding bottlenecks, and iterating through optimizations.
- Excited about agentic AI development workflows and using AI tools for automation.
- Strong technical collaboration skills across compiler, runtime, systems software, hardware, and performance modeling teams.
- Master’s or PhD in Computer Engineering, Electrical Engineering, Computer Science, or related field with strong experience in performance optimization, computer architecture, compilers, or systems software.
- Hands-on experience with runtime or compiler optimization (e.g., OpenJDK/JIT, LLVM, GCC, V8, Python).
- Strong understanding of CPU performance, memory hierarchies, concurrency, garbage collection, vector/SIMD optimization, and RISC-V architecture.
- Expertise with performance profiling and analysis tools (e.g., Linux perf, runtime profilers, QEMU, tracing tools).
- Strong programming skills in Java, Python, C/C++, and RISC-V assembly.
- Ability to work effectively across multiple software layers.
Benefits
- Highly competitive compensation package
- Benefits package