Overview As a Performance Architect on Tenstorrent's Platform Architecture team, you will bridge software execution and silicon design to optimize high-performance systems. You'll influence NoC, cache coherency, and interconnect decisions across ML software stacks, compilers, and CPU design. Your work helps shape the performance of our SoCs for AI and HPC workloads. This remote role places you in the US and offers a path to impact cutting-edge hardware and software co-design.
Compensation / Benefits- competitive compensation package
- remote work
Responsibilities- Collaborate across ML software stacks, compilers, CPU design, cache coherency protocols, and interconnect fabrics to shape high-performance systems
- Develop data-driven performance models and analyses to guide architectural decisions
- Bridge software execution and silicon design to optimize SoC performance
- Work with cross-functional teams to balance performance, complexity, and scalability
- Translate workloads into architectural insights that influence microarchitecture and interconnect decisions
Key requirements- BS/MS/PhD in EE, ECE, CE, or CS
- Deep understanding of NoC topologies, routing algorithms, QoS, and traffic scheduling
- Expertise in cache coherency protocols (AMBA CHI/AXI) and memory/IO technologies (DDR, LPDDR, GDDR, PCIe, CCIX, CXL)
- Proficiency in C/C++ and building efficient performance models
- Familiarity with ML/AI traffic patterns or formal verification of cache coherence protocols is a strong plus
- Collaborative mindset
- Analytical and data-driven
- Cross-functional communication
- NoC topologies and routing
- Cache coherency protocols (AMBA CHI/AXI)
- Memory/IO technologies (DDR, LPDDR, GDDR, PCIe, CCIX, CXL)