Cubic Corp. announced that its Cubic Mission Solutions (CMS) business division has launched a highly rugged, small form factor graphics processing unit (GPU) module designed to enable artificial intelligence and high-performance analytics at the extreme tactical edge.
Powered by dual NVIDIA GPU’s, the M3-SE-GPU delivers class-leading performance in a low power, compact and easily transportable solution. High throughput Ethernet connections combined with the dual processors enable parallel processing of distinct data streams, boosting throughput and allowing basic clustering operations.
The integrated deep learning engine and support for the NVIDIA JetPack SDK make the module ideal for artificial intelligence, machine learning applications and neural network development.
Benefiting from the M3-SE modular design, the M3-SE-GPU shares the same form factor and pass-through power connectors, allowing users to upgrade their current M3-SE server and network deployments with minimal changes to system design. This modular approach delivers great flexibility, as users can easily add GPU capabilities based on mission requirements, without the need to replace existing server or compute hardware.
Commercial GPU capability, flexible configurations and system commonality with the M3-SE family make the M3-SE-GPU the perfect complement to the compute platforms Cubic already provides to customers.
“Edge devices and sensors are generating massive amounts of data that must be processed in real time at the edge to be more relevant to current mission timelines,” said Mike Barthlow, senior vice president and general manager of rugged Internet of Things, Cubic Mission Solutions. “The new M3-SE-GPU module enables improved real-time intelligence at the edge, where and when it is needed most.”