S
SiMa.ai builds custom silicon and embedded software for real-time AI workloads in physical systems. The tech stack is deep-stacked in hardware design (SystemVerilog, UVM, Verilog, RTL logic) paired with ML inference frameworks (ONNX, TensorFlow, PyTorch) and embedded Linux ecosystems — a rare combination that signals end-to-end chip and software ownership. The hiring shape is almost entirely engineering (18 of 19 roles), with heavy concentration in principal and senior levels, indicating they're scaling a manufacturing and silicon-design operation rather than building a software-only product.
Tech Stack 49 technologies
Core StackPython C++ TensorFlow PyTorch UVM SystemVerilog C/C++ GStreamer ONNX ARM USB RISC-V Xilinx Versal OpenCV OpenCL MIPI CSI-2 V4L2 NVIDIA Jetson NXP i.MX Snapdragon Debian Yocto MIPI CSI Perl Tcl TCL LPDDR5 Verilog AXI Embedded Linux+19 more
What SiMa.ai Is Building
◆Challenges
- Accelerating hardware bring-up
- Ensuring robust manufacturing quality
- Reducing time-to-market
- Designing integrated hardware-software solutions
- Chip down design support
- Building end-to-end ai applications
- Optimizing real-time neural network algorithms
- Ppa optimization
- Timing area power constraints
- Lvs/drc violations
▲Active Projects
- Custom rtl logic development for interface optimization
- Pcie bring-up and debug on emulator
- Verification methodology for sima.ai’s mlsoc
- Gstreamer algorithm plugins
- Ddr bring up on mlsoc silicon
- Board test strategy for high-speed pcbas
- Clock network guidelines
- Solution porting
- End-to-end application development
- Real-time graph neural network algorithms for belle ii tracking detectors
Hiring Activity
20 roles · 8 in 30d
Department
Seniority
Company intelligence
