Nvidia faces competition for AI processing on smartphones

Nvidia is working to maintain its leadership in artificial intelligence as competition heats up. Major firms like Qualcomm and Micron are pushing for AI processes to move from data centers to devices like smartphones. Currently, most AI tasks, particularly "inference," occur in data centers, where Nvidia's chips hold a major share. As AI technology develops, Qualcomm argues that moving inference to smartphones can offer better availability, faster responses, and enhanced privacy. Qualcomm's Chief Financial Officer believes this shift is inevitable. However, there are challenges. Running AI on mobile devices can drain battery life, and existing chips may struggle with data speeds, creating delays for users. Micron is developing memory chips that aim to save power and improve performance for AI tasks. While Qualcomm and Micron hope consumers will choose devices equipped with their chips, smartphone sales are growing slowly. According to industry forecasts, shipments will only increase by about 2.3% this year. Nvidia, on the other hand, is proposing its chips to telecom companies. The company believes that integrating AI capabilities into local wireless networks can reduce lag and make better use of existing power. Despite concerns from telecoms about past investments in 5G, Nvidia has secured partnerships, including one with Samsung. Some industry experts are skeptical about the effectiveness of wireless networks for AI processing. However, Nvidia's current advantage remains strong, as many tech companies are invested in its chips for both training and inference tasks. The way these companies balance AI processing across data centers, wireless networks, and devices will shape the future competition.