본문 바로가기
bar_progress

Text Size

Close

Growing Expectations for Nvidia 'RTX 50'... MS CEO Calls It a "Game Changer"

New GPU Unveiled at CES 2025
Double the Computing Power
Lower Price, Higher Cost-Effectiveness

Jensen Huang, CEO of NVIDIA, has raised industry expectations for the newly unveiled graphics processing unit (GPU), the 'GeForce RTX 50 Series,' which was first revealed during his keynote speech at the world's largest electronics and IT exhibition, CES 2025. Influential figures leading the artificial intelligence (AI) era have also expressed interest, closely monitoring the potential impact it may have on the market.


Growing Expectations for Nvidia 'RTX 50'... MS CEO Calls It a "Game Changer" Jensen Huang, CEO of NVIDIA, is showcasing the GeForce RTX 5090 reference product during his keynote speech at CES 2025 on the 6th. Photo by AFP Yonhap News

According to industry sources on the 14th, Satya Nadella, CEO of Microsoft (MS), posted on his social networking service (SNS) LinkedIn account on the 10th, "Seeing the new RTX 50 Series GPU and the 'Windows Subsystem for Linux' (WSL) running NVIDIA NIM and Azure AI Foundry models on Windows 11 PCs is truly fascinating... This will be a game changer for running AI in edge computing environments."


The day marked the closing of CES 2025. After reviewing various new technologies presented at CES 2025, Nadella is interpreted to have highly evaluated the potential of the GeForce RTX 50 Series while summarizing key elements to watch on the exhibition's final day.


RTX is a GPU created by NVIDIA. Among them, the 5090 class is mainly used in high-end PCs such as workstations. It is utilized not only by gamers, the primary users of personal GPUs, but also in financial calculations, graphic computations, and scientific calculations, as well as extensively for AI research. On the other hand, NVIDIA's representative GPUs such as the V100, A100, and H100 are used in data center-grade servers for large-scale AI training.


Regarding the newly unveiled RTX 5090, CEO Huang explained that its data transfer rate is 1.8TB (terabytes) per second, which is twice that of the previous model, the RTX 4090. Huang emphasized that the lowest tier in the 50 Series, the 5070, delivers performance comparable to the previous generation's top model, the 4090, but at a significantly reduced price of $549. The 4090 was priced in the $1,500 range. NVIDIA has been releasing GPUs under the 'RTX' name since the RTX 2080 launched in September 2018, replacing the previous 'GeForce GTX' branding. The series has evolved through the 20, 30, and 40 Series to the current 50 Series.


There is also growing interest in whether Samsung Electronics will benefit from this. In his keynote speech, CEO Huang initially stated that the RTX 50 Series used Micron's graphics memory (GDDR7) products, but a day later officially corrected this by saying, "Samsung Electronics is supplying the initial batch," confirming the use of Samsung's GDDR7. This slip revealed internal information, and as the RTX 50 Series gains market success, attention toward Samsung Electronics' GDDR7 is expected to increase.


© The Asia Business Daily(www.asiae.co.kr). All rights reserved.

Special Coverage


Join us on social!

Top