ByteDance
System Architect - Server Hardware
San Jose, California · Full-time · Posted Oct 28, 2026
- Experience
- The posting states 3 years for each of two preferred experience areas; no minimum years are specified.
- Education
- Master’s degree or higher in Electrical Engineering, Computer Engineering, Computer Science, or related majors.
- Visa sponsorship
- International travel up to four times per year, including China, Europe, and South Asia; candidates must have a valid passport and be able to obtain necessary visas.
About the job
Architects and evaluates server and storage systems, integrating new technologies to improve performance, cost, and operability at scale.
What you'll do
- Evaluate emerging technologies through proof-of-concept systems and prototypes, and integrate suitable technologies into infrastructure.
- Develop server and hardware system designs with engineering, application, network, and data center teams to address workload needs and balance cost, performance, and operability.
- Engage with industry standards groups and technology partners to assess and contribute to emerging standards and designs.
What they ask for
- Master’s degree or higher in Electrical Engineering, Computer Engineering, Computer Science, or a related major.
- Deep understanding of computer system architecture.
- Knowledge of design principles and applications for CPU, memory, SSD, and network components.
- Understanding of virtualization, databases, distributed storage systems, and deep learning architecture; demonstrated cross-functional collaboration.
Nice to have
- Experience in hardware/software performance and power optimization.
- 3 years of experience in circuit board design, system debugging, power/signal integrity, and thermal solutions; or 3 years in server or storage system architecture design or software-hardware co-design.
Tools
CPU, memory, SSD, virtualization technology, database, distributed storage system, deep learning architecture
AI focus
The role evaluates and supports infrastructure for AI applications and deep learning workloads.
Model-written summary · the original posting is authoritative