Central to our mission is a highly programmable chip with workload specific hardware accelerators for compute, networking and storage workloads. Your role will be to define metrics that are interesting / salient for our system, implement software that gathers actual numbers for these metrics, profile scale out applications for performance bottleneck and analyze results.
Skills, Education, and Experience Required
- BS/MS in Computer Science or equivalent degree with 10+ years of software development
- Understanding of CPU architecture, system architecture and scale-out cluster architecture
- Strong algorithmic and data structure background
- Architectural sense: performance vs. storage tradeoffs, etc
- Adept in C or C++ programming
- Desire to push the state of the art
- Self-motivated, independent and proactive
Additional Success Factors
- Experience with defining benchmarks at sub-system level and for complete system
- Experience with networking and/or storage stack
- Fast development speed
- Startup experience