d-Matrix jobs, judged for you, free

What they do

d-Matrix designs chips and systems that run AI inference, the step where a trained model answers requests. Its design puts compute inside memory, to make inference faster and use less energy. Its Corsair card fits a standard PCIe slot and runs alone or beside GPUs. The Aviator SDK programs it. SquadRack and Squadron scale it to racks and data centers.

What they value

• Respect, collaboration, and direct communication: the culture rests on all three.
• Humility: the company values humble expertise.
• Different views: different perspectives lead to better solutions.
• Execution: the company wants people driven by execution and hard problems.
• Learning: continuous learning is part of the job.

Where they hire

Most roles are in Santa Clara, California, the headquarters. Most of them are hybrid, with three or more days a week in the office. A few roles are open to remote work in the United States. Single roles are in Bangalore, Belgrade, Sydney, Seattle, San Francisco, and New York. The postings say nothing about visa sponsorship.

Teams and roles

The founders previously built businesses that shipped over 100 million chips. Investors include Playground Global, M12, and Temasek. The company has offices in Silicon Valley, Sydney, Bengaluru, Toronto, and Belgrade. The board lists 37 roles.

• Software engineering: 14 roles in kernels, the SDK and compiler stack, inference execution, runtime, and quality assurance.
• Hardware engineering: analog and SerDes (high-speed serial link) circuit design, design verification, thermal design, and power and performance architecture.
• Architecture: a performance modeling architect, an LLM inference engineer, and an ML researcher. The researcher makes models smaller, through quantization and model compression.
• Systems engineering: PCB design and hardware qualification.
• General and administrative: site reliability engineering for AI infrastructure, information security, HR, recruiting, and investor relations.
• Sales, program management, and quality and reliability for AI hardware.

Two roles are contracts: a lab technician in Santa Clara and a recruiter in Bangalore. Most postings in the United States give a pay range, with equity and a bonus.

What they look for

Most roles are senior, at staff level or above. Postings ask for long industry experience, often 8 to 12 years or more with a bachelor's degree. They ask for fewer years with a master's degree or PhD. Degrees in electrical engineering, computer science, math, or physics are common. Software roles ask for C++ and Python on Linux. Kernel and compiler roles want PyTorch and compilers such as MLIR, LLVM, or TVM. Infrastructure and test roles ask for Kubernetes. Hardware roles ask for chip and board experience, such as SystemVerilog and UVM for verification.

Two test and quality postings want people who already use AI coding tools in their work.