On July 29, NVIDIA officially launched the Nvidia NIM platform, which combines AI models with accelerated computing resources and operates on the NVIDIA DGX Cloud platform. This platform aims to simplify the deployment process of generative AI and is designed specifically for large-scale applications.

NVIDIA has updated its Nvidia AI Enterprise products with a series of enhancements, including the addition of the cloud-native microservice NIM's inference microservice library, which covers physical world environments and advanced visual modeling. These software and services require NVIDIA hardware to operate, costing $4,500 per GPU annually.

NVIDIA has also partnered with Getty Images (GETY) and Shutterstock (SSTK) to ensure that AI-generated images more accurately match textual prompts. Additionally, NVIDIA's software and services enable users of Apple's Vision Pro headset to create virtual worlds.

Over the past six months, NVIDIA's stock price has risen by 78.63%, contributing over 35% to the market capitalization growth of the S&P 500 index. However, NVIDIA has recently significantly reduced the supply of high-end graphics cards (GPUs), which may be due to tight AI chip supplies and preparations for the next-generation RTX 50 series graphics cards. The market expects that the supply of RTX 4070 and higher models will become even tighter in August.

With NVIDIA graphics card prices rising, Taiwanese graphics card manufacturers such as ASUS, MSI, and Gigabyte are expected to benefit. This is particularly true against the backdrop of growing demand for AI servers, which has led to simultaneous increases in graphics card prices, boosting the market outlook for related Taiwanese manufacturers.

Industry forecasts predict that with the launch of the RTX 50 series graphics cards by the end of the year, the shipment volume and average selling price (ASP) of graphics cards will increase.