Overview In this Architect role, you will shape the silicon architecture for high-performance, sustainable GenAI systems. You will define requirements, guide full-chip and subsystem design, and collaborate with Research, Software, and Hardware teams to deliver performant hardware across the product lifecycle. You'll translate workloads and ML research into architectural choices that balance area, timing, and power while interfacing with vendors and customers. This is a chance to impact how large AI models are trained and run at scale, contributing to MatX's end-to-end stack.
Compensation / Benefits- 4 weeks PTO (accrued)
- health insurance for employees and dependents
- retirement plans with company contribution
- paid parental leave
- professional development budget
- commute reimbursement
Responsibilities- Own hardware architecture across the product lifecycle
- Understand customer workloads and ML system research to guide design
- Produce architecture requirements and specifications
- Estimate area, timing, and power to inform trade-offs
- Collaborate with Research, Software, and Hardware teams
- Interface with vendors and customers
- Parallel Compute: SIMD, Vector, VLIW models in architecture
- Data Movement: efficient storage and communication across memory and networks
- Distributed Systems: workload mapping, bottleneck analysis, scalable systems
Key requirements- Expertise in modern high-performance CPU, GPU, or AI accelerator architecture
- Experience analyzing and optimizing system performance with diverse tools and methodologies
- Proficiency in programming and code optimization
- Excellent writing skills
- collaboration with cross-functional teams
- strong written communication
- problem-solving mindset
- high-performance CPU/GPU/AI accelerator architecture
- performance analysis and optimization
- parallel compute models (SIMD, Vector, VLIW)
