Recently, Nvidia released a series of chips focusing on high-end products, including Ada Lovelace RTX 40 series game GPUs, automatic driving chip Thor, etc., and announced that the high-performance data center GPU H100 was fully put into production.
Nvidia's RTX 40 series GPUs released this time switched to chip manufacturers, adopting the 4-nanometer process optimized by TSMC for GPUs, while the previous generation of RTX 30 series was produced by Samsung.
For consumers, the new generation of high-end game graphics cards are significantly more expensive than previous versions. Among them, Nvidia RTX 4090 is priced at US $1599, 7% higher than the previous generation 3090 graphics card sold in 2020, and is scheduled to be sold on October 12; The mid-level graphics card RTX4080 is priced at $899, 29% higher than the previous generation 3080 graphics card, which will be available in November.
Ada Lovelace of RTX 40 series mainly replaces the Ampere architecture, which was launched in May 2020, while RTX 30 series game graphics cards based on Ampere architecture were launched in September 2020.
According to the report of MoneyDJ on September 22, Nvidia's latest release of the next generation Ada Lovelace architecture graphics processor (GPU) triggered complaints from players because of the price hike. In this regard, Huang Renxun, CEO of Nvidia, said that the death of Moore's Law was the reason why the company had to raise its price.
Regarding the statement that the price is too high, Huang Renxun said that the increased price is reasonable because the Lovelace architecture can support Nvidia to enter the metaverse. The price of 12 inch silicon wafers today is a ton more expensive than in the past, not just a little expensive. Moore's Law is dead, and now it is impossible to deliver twice the efficiency at the same cost every year and a half. That is to say, "The era when chips can reduce costs every once in a while has ended." "Computing is not a problem of chips, but of software and chips," he said
In addition, Huang Renxun revealed that Nvidia is ready to work with Chinese customers to move the database of the meta universe software platform "Omniverse" to the cloud. Nvidia and Deloitte announced on September 22 to expand cooperation to help global enterprises develop, implement and deploy hybrid cloud solutions using Nvidia Artificial Intelligence (AI) and Nvidia Omniverse Enterprise platforms.
The expanded cooperation between the two sides will help global enterprises to more easily establish and operate the most advanced AI and meta universe services, including Internet of Things edge AI, voice AI, recommendation systems, customer service chat robots, network security and digital twins.
Source of cover picture: Paixin