Demos include AI Processing of 100 simultaneous video streams, industrial defect detection, Yolo-World, and a comparison with NVIDIA GPUs
ANN ARBOR, Mich., Dec. 19, 2024 /PRNewswire/ -- MemryX Inc., a leader in edge AI accelerators, will unveil cutting-edge demonstrations of the MX3 AI Accelerator during the 2025 Consumer Electronics Show (CES) in Las Vegas. Taking place at the Palazzo Hotel from Jan. 7–10, demos will highlight the MX3 industry leading performance, efficiency, and versatility in real-world Edge AI applications.
Featured demos include:
- 100 Simultaneous Video Streams: AI-enabled video management system for advanced surveillance, safety, and analytics applications.
- Industry 4.0: AI-driven industrial manufacturing quality inspection system demonstration together with DYNICS hardware and Telit software.
- Yolo-World: High performance real-time processing using the latest cutting edge "zero-shot" object detection model that uses open-vocabulary detection.
- GPU Comparison: MemryX MX3 AI Accelerator competes side-by-side with mainstream NVIDIA GPU, showcasing the MX3 Edge AI performance advantage while using just a fraction of the power.
In Q4 2024, MemryX announced the availability of production MX3 modules and the public release of a Developer Hub. This, along with a recent online developer community offers an full ecosystem of tools and resources to simplify AI implementation and accelerate deployment timelines.
"CES 2025 provides the first opportunity for us to showcase MemryX production quality hardware and software already in the hands of many customers," said Keith Kressin, CEO of MemryX. "We are excited to showcase our unique combination of efficiency, scalability, and ease of use that enables customers to quickly and efficiently deploy AI solutions."
"ASUS has integrated the MemryX MX3 into our edge AI devices, and the results have been transformative," said Jessy Li AIoT Functional Director, Smart Manufacturing, ASUS. "The seamless integration and powerful performance of the MX3 have opened new possibilities for AI at the edge, delivering an exceptional experience for our customers."
"Our collaboration with MemryX brings cutting-edge AI acceleration to industrial automation, as showcased in our Xi-MX3 system performing real-time quality inspections," said Ed Gatt, DYNICS President. "These types of integrated solutions not only streamline operational efficiency but also empower businesses to fully embrace the transformative potential of Industry 4.0, driving innovation, productivity, and a competitive edge in the evolving industrial landscape."
MemryX's MX3 AI Accelerator is built to deliver unparalleled versatility and performance for edge AI applications. Key benefits include:
Best-in-class performance-per-watt: Optimized for low power consumption without compromising on speed or accuracy.
Broad compatibility: Supports popular AI frameworks (e.g., TensorFlow, ONNX, PyTorch), host processors (e.g., x86, Arm, RISC-V), and operating systems (e.g., Linux, Windows, Android).
Scalability and reliability: Handles complex AI workloads with ease, from consumer electronics to industrial automation.
MemryX invites CES attendees to experience the groundbreaking demos and connect with its leadership team. To schedule a meeting, email alan.frost@memryx.com. Walk-ins are welcome at the MemryX suite at the Palazzo Hotel during CES.
About MemryX
MemryX Inc. is a fabless semiconductor company pioneering AI processing solutions for edge devices. Powered by proprietary compute-at-memory technology, MemryX accelerators combine high performance and low power to enable efficient, scalable AI for industries including IoT, transportation, and industrial automation. MemryX is backed by top investors and has headquarters in Ann Arbor, Michigan, with offices in Bangalore, India, and Taipei and Hsinchu, Taiwan. Visit MemryX.
**Media Contact**
Alan Frost
Head of Corporate Marketing, MemryX
Email: alan.frost@memryx.com
This updated release combines CEO and customer quotes to underscore MemryX's innovation and real-world impact.
SOURCE MemryX
