New Grad Software Engineer

Position:
New Grad Software Engineer — Physical AI Compute

Xcelerium is building the compute foundation for real-time Physical AI — intelligent systems that must process rich sensor data, make decisions, and act under demanding real‑world constraints of latency, bandwidth, power, size, and reliability.

Our X²PU architecture is a unified polymorphic compute platform designed for workloads that span signal processing, AI/ML, linear algebra, optimization, control, and real‑time decision‑making. Xcelerium is developing this platform through full‑stack hardware, software, and algorithm co‑design for applications such as autonomous systems, robotics, wireless infrastructure, aerospace and defense, industrial automation, and edge AI.

Xcelerium is led by an experienced team of computer architects, chip designers, software leaders, and semiconductor executives with deep backgrounds in advanced SoCs, wireless, AI/ML, compute, and high‑volume technology products. This is an opportunity to join a team building a new class of compute platform for real‑world AI.

Primary location:
Irvine, CA
Alternate location:
Santa Clara, CA
This is an on‑site role.

Full‑time, new graduate / entry‑level

Expected pay: 80,000‑100,000 USD per year
Additional compensation:
Bonus Equity package

• Medical
• Dental
• Paid time off
• 401(k) match
• Vision

As a New Grad Software Engineer — Physical AI Compute, you will help build the software foundation for Xcelerium’s X²PU architecture.

This role sits at the intersection of computer architecture, compilers, SDKs, performance modeling, AI frameworks, and Physical AI applications. You will contribute to tools and software that allow developers to understand, program, optimize, and deploy workloads on a new X²PU architecture purpose‑built for real‑time Physical AI.

You will work with leading computer architects and expert chip designers to help define the software stack for a new class of real‑time compute.

We are looking for highly curious, technically strong, and ambitious engineers with excellent programming skills, strong systems instincts, and the desire to learn quickly in a deep technical environment.

• Build architectural models, simulators, and performance‑analysis tools for the X²PU architecture.
• Contribute to compiler, runtime, SDK, and developer‑tooling infrastructure for programming and optimizing X²PU workloads.
• Help map Physical AI workloads onto a new computer architecture, including signal processing, AI/ML, linear algebra, optimization, and control workloads.
• Develop and optimize software kernels, libraries, APIs, examples, and application demos.
• Work on software tools that help evaluate latency, throughput, memory movement, utilization, power, and performance tradeoffs.
• Collaborate with architects, chip designers, firmware engineers, and application developers to connect hardware capabilities with software usability.
• Prototype Physical AI applications in areas such as sensing, perception, autonomy, robotics, wireless, or real‑time edge AI.
• Integrate with or build around modern AI and compiler ecosystems where appropriate.
• Write clean, well‑tested, well‑documented code that can grow into production‑quality software infrastructure.
• Participate in technical design reviews and help shape how developers program a new X²PU platform.

• B.S. or M.S. in Computer Science, Computer Engineering, Electrical Engineering, or a related field, or equivalent hands‑on project/internship experience.
• Strong programming skills in C/C++ and Python.
• Strong fundamentals in data structures, algorithms, systems programming, debugging, and software engineering.
• Familiarity with computer architecture concepts such as memory hierarchy, parallelism, instruction execution, vector/SIMD processing, accelerators, GPUs, DSPs, or embedded systems.
• Familiarity with compilers, runtimes, SDKs, AI frameworks, or performance modeling.
• Comfort working in Linux development environments using tools such as Git, build systems, debuggers, profilers, and test frameworks.
• Strong analytical ability and interest in performance, efficiency, latency, and…