DeepX Lays Groundwork for Public Market Push With Low-Power Inference Certification
AI TIMES ·
✦ AI Summary
According to AI TIMES, DeepX said on September 14 that it received the government’s green technology certification for its low-power AI infe…
According to AI TIMES, DeepX said on September 14 that it received the government’s green technology certification for its low-power AI inference technology for edge computing, and that its related AI accelerator module also obtained green technology product verification. In a certified test, the company said the key basis for the certification was that the system cut power consumption by up to 26% compared with a reference system and reduced operating temperatures by more than 20 degrees. The company explained that the certification was built on its in-house NPU design, which reduces data movement and memory access while maintaining computing performance. The verification also covered both products for small edge devices and products for high-performance inference, underscoring the broad scope of their application. Green certification is a system under which the government officially recognizes technologies related to climate response and new industries, giving it significance not only for product performance but also for commercialization and market entry. DeepX expects the certification to help with public procurement and expand its business in South Korea and abroad.
Perspective
The significance of this case is that low-power performance has moved beyond simple technology promotion and now serves as a qualification that improves market access. In particular, the confirmation of both power savings and heat reduction sends a signal that it can ease the operational burden long seen as an obstacle to the spread of on-device AI. In the end, competitiveness is likely to be judged not just by how fast a system is, but by how stably it runs on how little power.
This perspective is BizCrush's own commentary and is not part of the reporting by AI TIMES.
This article was produced with the help of an automated content generation algorithm.
Source: AI TIMES
View originalThis article was summarized and organized by BizCrush based on the original article from AI TIMES. For exact quotations and full details, please refer to the original article.