MatX
Developing custom silicon chips optimized for large language models.
Company facts
- Founded
- 2023
- Employees
- 100
- Headquarters
- United States
- Type
- Private
- Funding
- Series B · $500M · 2026
Open positions
Software Engineer - Kernels
Designs and optimizes compute kernels that interface directly with MatX's custom hardware for training and running the largest ML/AGI workloads, working alongside ML research and hardware teams. Stack centers on Rust plus low-level optimization (C/C++, assembly, SIMD, GPU/CUDA); hybrid role, 3 days/week in Mountain View.
SOC Integration Engineer
A silicon SOC integration engineer who stitches subsystems into the full-chip top-level RTL for MatX's AI training chips, building automation for on-the-fly SOC RTL generation and aligning clock/reset, timing, and floorplan connectivity with physical design. Core tools: SystemVerilog, Python/Perl, and SOC integration/verification flows.
Software Engineer, AI Tooling
Builds the AI agent infrastructure at MatX, a custom AI silicon company: making coding agents first-class users of internal infra with their own identities and audited access, running them in ephemeral VMs and no-egress sandboxes, and maintaining a company knowledge base pulled from the monorepo, design docs, and Slack.
Lead Software Engineer, Kernels
MatX is hiring a hands-on tech lead to manage and grow a kernel engineering team while personally designing high-performance compute kernels for its custom AI hardware. Day to day blends people leadership with low-level optimization work (C/C++, assembly, SIMD, GPU/CUDA) in close partnership with ML research and hardware teams.
System Software Engineer, Linux Driver and Driver Adjacent
Build the Linux kernel driver stack (PCIe, memory management, DMA, interrupts) and low-level userspace libraries for MatX's AI accelerator silicon, from pre-silicon simulation/emulation/FPGA testing through first-silicon bring-up and performance tuning. Core tech: C, Linux kernel, PCIe, hardware debugging.