DeepX Launches Low-Power M1 Chipset Dedicated to Large-Scale Video Processing
A New Philosophy Beyond Just a Cheaper GPU Alternative
Proven Technological Excellence with Innovation Award at CES 2026
Artificial intelligence (AI) semiconductor company DeepX has launched a large-scale, low-power chipset dedicated to video processing.
On December 8, DeepX announced the release of its video intelligence dedicated chipset, the 'DX-H1 V-NPU,' which can process AI analysis for hundreds of video channels at a low power consumption level of around 30W. This product integrates video input, compression, and AI inference processes-which were previously divided between graphics processing unit (GPU) servers and separate codec equipment-into a single card. As a result, it can reduce the power consumption, cost, and complexity involved in large-scale video AI processing.
DeepX has launched the video intelligence dedicated chipset 'DX-H1 V-NPU'. DeepX
View original imageDeepX emphasized that the DX-H1 V-NPU will serve as a turning point, shifting the basic unit of video AI infrastructure from the GPU to the neural processing unit (NPU). Compared to GPUs, this product reduces hardware costs by approximately 80% and power costs by about 85% for the same number of channels, while maintaining 24-hour real-time inference performance.
This is especially significant as it provides a structural solution to support the growing demand for video AI in urban and industrial sites in a sustainable way, amid increasing power shortages and uncertainties in GPU supply for data centers.
DeepX views this not simply as a "cheaper alternative" to GPUs, but as a product that fundamentally redefines the design philosophy of video intelligence infrastructure. While GPUs excel at general-purpose computation, they are not optimized for multi-channel video input/output and real-time streaming. In contrast, the DX-H1 V-NPU has been optimized-through its memory hierarchy and other features-for environments where video streams are generated every second rather than every minute, enabling seamless channel intake, AI model inference, and re-compression for output.
A DeepX representative explained, "Large-scale video AI will no longer rely on surplus resources from general-purpose GPU computation, but will instead become 'an industry of its own' running on dedicated chipsets." The company aims to make the NPU the core of the entire server architecture in areas such as smart cities, traffic control centers, and national infrastructure.
Hot Picks Today
"Two Spend 500,000 Won in a Flash, Yet Wait Two Hours?" ... 'Dreamland' Shunned by Kids and Foreign Tourists
- The Era of Getting 200 Million Won After 30 Years Is Over... A Japanese Bank's "Perpetual Department Head" Who Just Clocked In Cries Out
- Spain Hit by Its Worst Tragedy... Lee Kangin, Moreno and Real Madrid Join Soccer World in Mourning
- They Said, “Make It Your Home for Life” - Then Suddenly Kicked 200 Passengers Off... What Happened to the Around-the-World Cruise?
- "130 Rice Bowls Sold Out in a Day"... 'Unfiltered Face' Adds 8 Million Won to Night Market Sales: China's Bungeunwoo
The DX-H1 V-NPU will be officially unveiled at CES 2026, the world’s largest technology exhibition, to be held in Las Vegas, USA in January next year, where it has already proven its technological prowess by winning an Innovation Award.
© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.