On September 21, chip leader Nvidia announced the launch of the latest "GeForce RTX 40" graphics processing unit (GPU). NVIDIA emphasized that the new flagship product RTX 4090 is the fastest gaming GPU on the surface.
The legal person is optimistic that NVIDIA's new orders will be bagged, which will assist TSMC in receiving orders for advanced manufacturing processes, and NVIDIA's new product performance is the highest in the industry, which further highlights TSMC's technical capabilities in advanced manufacturing processes.
The market has high expectations for NVIDIA's new chips. The industry originally expected it to be released in August. However, due to the high inventory level of graphics cards, NVIDIA's new products have been "only heard the sound of the stairs", and they were officially released yesterday.
Huida pointed out that the RTX 4090 will be launched on October 12, priced from NT$56,990. In addition, the "RTX 4080" will also be launched, and it will be available in 16GB and 12GB versions. It will be available in November. , 16GB starts at 42,990 yuan, and 12GB starts at 31,990 yuan.
According to the current status of the discrete graphics card market, NVIDIA has a market share of 70-80%, far exceeding the 20% of its competitor Super Micro. In addition, Intel has also returned to the discrete graphics battlefield.
This time, NVIDIA's new graphics processor was transferred from Samsung to TSMC. Regarding the media inquiries about the cooperation between NVIDIA and Samsung, NVIDIA founder and CEO Huang Renxun said that Huida and Samsung's foundry department have long-term and close cooperation. Spanning several technology generations, Samsung also provides High Bandwidth Memory (HBM) to Huida.
Huida said that the RTX 40 series is the world's first GPU product using the new Ada Lovelace architecture, enabling real-time ray tracing and neural rendering using artificial intelligence (AI) to generate pixels into a new era.
Huang Renxun said in the live broadcast of the GeForce Beyond speech at the GTC conference that the era of RTX ray tracing and neural rendering is in full swing, and the new Ada Lovelace architecture improves the performance of ray tracing and neural rendering.
The Ada architecture is four times more efficient than its predecessor, setting a new standard in the GPU industry.