NVIDIA announced today that the mainstream driver support for Fermi architecture graphics cards has officially ended. No future features, performance optimizations, bug fixes will be provided, and security issues will only be urgently resolved when necessary, and only until 2019. January. The Fermi Architecture was born in April 2010 and is named after the famous Italian physicist Enrique Fermi. It is the successor to the Tesla architecture and the oldest NVIDIA GPU supporting DX12. It was initially manufactured on a 40nm process, and the later mobile version was upgraded to a 28nm process, integrating up to 3 billion transistors.
Fermi architecture is used for GeForce 400/500 series graphics cards on desktops and laptops. Some of the cores even continue into the GeForce 800 series in the form of a vest. A total of 72 products are derived, and Quadro x000 is also used. Series, Quadro NVS Series, Tesla Computing Cards.
In the future, NVIDIA’s official mainstream graphics architecture family includes Kepler, Maxwell, Pascal.
It is worth mentioning that both the Fermi and Tesla architectures supported for eight years, but Tesla’s extended security support lasted for two years, and Fermi only 10 months.
At the same time, starting this month, NVIDIA is no longer mainstream support for 32-bit systems, also limited to emergency security repairs, as of January 2019. GeForce Experience software will not be updated on 32-bit systems.
