Shulou(Shulou.com)11/24 Report--

Thanks to CTOnews.com netizen OC_Formula and soft media Xinyou 1933769 for the clue delivery! CTOnews.com, April 9 (Xinhua)-- Qualcomm and Nvidia, the two leading manufacturers of artificial intelligence chips, beat Nvidia 2-1 in three measures of power efficiency in a new set of test data released Wednesday.

According to CTOnews.com, artificial intelligence models need to be trained with a large amount of data to improve their accuracy and performance. After the training, the artificial intelligence model can be used for reasoning, that is, to perform specific tasks, such as generating a text reply based on input, or determining whether a picture contains a cat. Reasoning is a link in which artificial intelligence technology is widely used in products, but it will also increase the cost of enterprises, one of the main costs is electricity.

Drawing on its experience in designing chips for low-power devices such as mobile phones, Qualcomm has launched a chip called Cloud AI 100, which provides high-performance, low-power artificial intelligence processing for cloud and edge. The chip beat Nvidia's flagship chip H100 on two power efficiency indicators in test data released on Wednesday by MLCommons, an engineering alliance that maintains testing standards for the artificial intelligence chip industry.

The power efficiency index refers to how many server queries can be performed per watt of power. Qualcomm's Cloud AI 100 achieves 227.4 queries per watt for image classification, while the Nvidia H100 has only 108.4 queries per watt. Image classification can be used to identify objects or scenes in pictures. Qualcomm also leads Nvidia in object detection, with 3.8 queries per watt and 2.4 queries per watt, respectively. Object detection can be used to analyze surveillance videos of retail stores to find out where customers go most often.

However, Nvidia has an absolute advantage in natural language processing, ranking first in terms of performance and power efficiency. Natural language processing is the most widely used artificial intelligence technology in systems such as chat robots. Nvidia achieved 10.8 queries per watt, while Qualcomm ranked second with 8.9 queries per watt.

Both Qualcomm and Nvidia hope to capture data center market share by providing efficient artificial intelligence chips. As more and more enterprises integrate artificial intelligence technology into their products, the market is expected to grow rapidly.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.