Software IP for video codecs, compilers, and autonomous vehicle perception
MulticoreWare builds compiler toolchains, video codec libraries, and sensor-fusion analytics for automotive (ADAS/autonomous), surveillance, and industrial applications. The tech stack is infrastructure-heavy—Kubernetes, OpenStack, Docker, Terraform, AWS/Azure/GCP—suggesting a services-and-platform hybrid model. Active projects focus on container orchestration and compiler IR optimization, indicating sustained backend infrastructure work, though hiring velocity is minimal (no new roles posted in 30 days) and skewed toward senior engineers and one intern in India.
Founded in 2009, MulticoreWare develops software IP and engineering services for companies building perception systems and video pipelines. The product portfolio spans three categories: compiler toolchains and code generation for heterogeneous hardware (CPUs, GPUs, specialized accelerators); video codec libraries (x266, x265, Ultraziq) deployed in broadcast and streaming services; and AI analytics solutions using camera, radar, LiDAR, and IMU sensor fusion for ADAS and autonomous-vehicle applications. Customers span automotive OEMs, surveillance platforms, defense, medical imaging, IoT, and robotics. The company operates at 201–500 employees, headquartered in San Jose, CA, with current engineering focus on cloud-native container platforms and compiler backend optimization.
Core infrastructure: Kubernetes, OpenStack, Docker, Terraform, AWS, Azure, GCP. Development: LLVM, Git, Linux. Specializations span heterogeneous computing, sensor fusion (radar, LiDAR, IMU), CUDA optimization, and neural networks.
Container orchestration with Kubernetes and OpenStack integration, cloud-native CI/CD pipelines, compiler backend code generation for custom hardware, and IR-level optimization passes for heterogeneous platforms.
Other companies in the same industry, closest in size