A global car manufacturer leveraged our complete portfolio of high-performance data loggers and edge computers for its ADAS and autonomous driving development. Those systems enabled AI inference and unprecedented data logging to enable the inference and training of fully autonomous vehicles during test phases.
REQUIREMENTS AND CHALLENGES
Logging speed, storage capacity, and ease of data transfer to lab servers are key requirements for the data logging phase, whereas AI processing performance and efficient cooling are mandatory for in-vehicle AI. Sensor data is often stored at full resolution to perform in-vehicle, real-time operations (AI inference or reinforcement training) or offline AI training or simulations in large lab workstations.
Moreover, the high-performance computers had to be ruggedized to endure the harsh environmental conditions of automotive applications and meet industry requirements with specific certifications.
They also needed to be installed into the car trunk, so space was very limited: they needed to be extremely compact, while sustaining server-class data processing, at the same time being able to dissipate the heat generated by the high-performance data processing.
WHAT WE PROVIDED
- Rugged, automotive-certified high-performance data loggers (up to 28 GB/s of sustained logging speed) and edge computers (up to 164 TFLOPS)
- Extended expertise in liquid cooling to provide maximum computational density employing fanless and ventless systems
- Ease of installation in the vehicle trunk and fast data transfer to customer’s servers
HOW THE CUSTOMER MET THEIR GOALS
The customer could rely on a complete portfolio of edge computing systems for every phase of its ADAS and autonomous driving application: from high-speed data loggers for fast data recording, to edge AI systems with powerful GPUs for the real-time AI inference in the vehicle, up to the data transfer to the customer’s servers for AI training and validation.