Software Engineer, Supercomputer Network (Starlink)

SpaceX•Palo Alto, CA

About The Position

SpaceX is developing technologies intended to enable human life on Mars and designs, builds, and operates large AI supercomputers. As a member of the Platform Core team, the Software Engineer is responsible for low-level software on devices used in the supercomputer and its network fabric. The role involves building and maintaining the Linux-based operating system and software that interfaces directly with the supercomputer’s hardware and network fabric, as well as the Starlink laser mesh of satellites. The position includes developing device drivers, supporting hardware bring-up, and managing platform deployment, operation, and debugging in production. It also involves working with hardware teams and external partners on future supercomputer hardware and bringing up the operating system and software stack on that hardware.

Requirements

Bachelor’s degree in computer science, engineering, mathematics, or a science discipline, or at least 2 years of professional software development experience in lieu of a degree; internship experience is applicable.
Hands-on system programming experience in C or C++.

Nice To Haves

Deep knowledge of the Linux kernel and its networking stack, or experience with another production-grade kernel.
Proficiency with debugging tools across kernel, network, and userspace, such as ftrace, perf, Wireshark/tcpdump, eBPF, and gdb.
Experience bringing up new hardware from scratch, such as flashing bare-metal software on a microcontroller with JTAG or bootstrapping a Linux kernel and root filesystem on new hardware.
A track record of developing and optimizing device or peripheral drivers, ideally for high-speed interfaces or with considerations such as DMA and cache coherency.
Understanding of peripheral hardware interface standards such as Ethernet, PCIe, I2C, and SPI; ability to critically review schematics and debug interfaces as needed.
Experience with product security, operating-system or system hardening, and secure boot schemes.

Responsibilities

Build and maintain a lean, high-reliability Linux-based operating system for the supercomputer network fabric and Starlink laser mesh of satellites.
Write or rewrite high-performance device drivers to maximize the capabilities of network and compute hardware.
Manage deployment, operation, and debugging of the platform in production.
Investigate and resolve anomalies, identify root causes, and improve system reliability.
Develop and continually improve observability and configuration-management tools.
Collaborate with hardware teams and external partners to design next-generation supercomputer hardware and bring up the operating system and software stack on it.

Benefits

Potential long-term incentives in the form of company stock or long-term cash awards.
Potential discretionary bonuses.
Ability to purchase additional stock at a discount through an Employee Stock Purchase Plan.
Comprehensive medical coverage.
Vision insurance.
Dental insurance.
401(k) retirement plan.
Short- and long-term disability insurance.
Life insurance.
Paid parental leave.
Various other discounts and perks.
May accrue 3 weeks of paid vacation.
Eligible for 10 or more paid holidays per year.
Washington State employees accrue paid sick time in compliance with state and federal law.
Company shuttles for roundtrip travel from select Seattle locations to the SpaceX Redmond office, Monday to Friday.