In the case of the GeForce RTX 4090, the graphics card uses the AD102-300-A1 chip modification, in which 16,384 CUDA cores are activated. However, the developers are keeping the full-featured version for future solutions. Returning to the AD102, it should be noted that in the maximum configuration, the GPU accommodates 18,432 CUDA cores. Technological progress in the manufacture of crystals can still surprise and allows to significant increase the capabilities of microcircuits. When you operate with such indicators, the fantastic difference becomes more visible. If we calculate the relative density of the composition of the elements, then it has increased almost three times – from 45.1 million transistors per 1 mm² to 125.5 million per 1 mm². For comparison, the top processor of the previous generation (Ampere), which was manufactured by Samsung according to the standards of 8 nm, had an area of 628 mm² and included 28.3 billion transistors. This is a fairly large GPU with a crystal area of 608 mm², which contains 76.3 billion transistors. The chips are manufactured by the Taiwanese company TSMC according to technological standards of 4 nm (TSMC 4N). GeForce RTX 4090 graphics cards are based on AD102 graphics processors with Ada Lovelace architecture.

In addition to the fact that this model deserves special attention, we start our first acquaintance with the adapters of the line with MSI GeForce RTX 4090 SUPRIM LIQUID X 24G – an original video card with a liquid cooling system. So it is not surprising that the developers began to reveal the RTX 40xx graphics line with the flagship of the series – GeForce RTX 4090 24 GB.

With each next generation of graphics cards, the NVIDIA company tries to impress so much that potential buyers think about another upgrade, even in cases where they own top models of the previous generation and did not plan to change anything in their configurations at all.

MSI GeForce RTX 4090 SUPRIM LIQUID X 24G.

Msi laptop fan speed control