Edge AI {hardware} producer Unigen has launched the Poptart E3.S AI Module, increasing its AI product portfolio with an industry-standard type issue for air-cooled AI inference servers.
The module integrates with AMD Genoa servers and Community Optix’s VMS AI software program, enabling environment friendly processing of lots of of video streams from IP safety cameras for functions like good cities, warehouses, and transportation techniques.
Poptart E3.S delivers 52 TOPS efficiency utilizing 10 Watts, powered by two Hailo-8 Edge AI processors, providing superior energy effectivity in comparison with different options.
It helps plug-and-play and hot-swap performance, becoming into E3.S slots sometimes used for SSDs, enhancing AI capabilities in massive server setups with decrease energy consumption than GPU modules.
The module helps a number of parallel neural networks and might perform as a unified Massive Language Mannequin array for advanced AI duties. It additionally comes with a sturdy software program suite supporting AI frameworks like TensorFlow, PyTorch, and ONNX, and features a dataflow compiler for simple neural community mannequin integration.
This announcement underscores the rising shift in the direction of specialised, energy-efficient {hardware} tailor-made particularly for edge AI workloads.
Unigen additionally just lately unveiled DDR5 reminiscence modules to satisfy rising calls for in AI and industrial functions.
Associated
AI/ML | AMD | edge AI | edge servers | Hailo | {hardware} | Unigen
