Nvidia trademark protector Hopper – follows the architecture Ampere

In 2020, Nvidia is expected to take the step down to 7 nanometers, which guarantees both higher performance and lower power consumption. The architecture in question will be Ampere, which is believed to be a shrinkage and refinement of Turing, which was launched in late summer 2018 together with the Geforce RTX 2000 series.

Some time ago, the first information about the next architecture was published, which according to the information would be named Hopper after the computer researcher Grace Hopper. Nvidia is now applying for trademark protection for Hopper in particular, which strengthens the information that it will be a code name for a future architecture.

According to previous information, Hopper means a radical change in how graphics circuits are designed. The architecture is said to take the step to the Multi-Chip Module (MCM), where several separate circuits are manufactured and joined together on a substrate to create a higher performing solution.

The solution itself is not new, but was used by Intel almost a decade ago with the Clarkdale processor family. During the summer, AMD breathed new life into the concept on the consumer side with the Ryzen 3000 series, where the company uses a central circuit for communication and two computing circuits to deliver up to 16 cores in a standard processor for socket AM4.

For Nvidia, such a solution would mean that the company can increase performance more than would otherwise be possible. In recent generations, Nvidia has made huge graphics circuits close to the limit (reticle limit) for what is possible with today’s lithographic processes.

Whether the name of the architecture after Ampere will be Hopper or whether it is a later story remains to be seen. First, Nvidia will launch Ampere, which will appear sometime in 2020.

Source: Videocardz

Read more about Nvidia’s upcoming architectures: