Capable of processing multiple streams in parallel, each with the same or different models.
New fully software programmable graph streaming architecture designed for edge applications and inference giving low latency.
Blaize Xplorer X1600P PCIe accelerator runs industry standard AI workloads, with 15 and 20 TOPS product running with and average power under 15W.
Blaize AI Studio and Picasso SDK is extremely flexible for building end-to-end applications, with a no/low code option for those new to AI through to a C/C++ option for the most capable developers.
Delivering AI solutions capable of operating 24/7, without interruption providing complete real-time processing and data analyses.
With AI the models can be improved and updated overtime to give better accuracy and more reliable results.
Edge AI can identify vehicles', makes, models, drivers and license plates for speed limit control, to avoid accidents, to monitor traffic volume on the road and control traffic lights preventing traffic build up and allowing people to avoid congested areas.
With AI and deep learning many smart city applications like access control, pedestrian and buyer behavior, retail analysis and more are based body pose and actions can be developed. This will lead to more efficient and safer systems for people.
From material inspection, production-line automation to safety systems in manufacturing facilities, machine learning allows factory operation to sense, act and respond in real-time based on desired outcomes and not programmed behavior to save costs and increase quality and throughput.