Today the embargo on mobile graphics of the GeForce RTX 4000 series ends. In today’s article, we will first test the RTX 4090 graphics itself. The performance of the processor and the test of the entire notebook will be in a separate test.
Just for the sake of time, let’s remember when the previous generations of Nvidia mobile graphics were released. The highest Ampere model presented by the RTX 3080 Mobile went on the market in January 2021. The graphics carried 6144 CUDA cores and has 8 GB of GDDR6 memory. Exactly one year later, Nvidia launched an improved version on the GA103 core. It is the current, most powerful mobile graphics RTX 3080 Ti (Mobile). It carries 7424 CUDA cores with 16 GB of GDDR6 memory.
Now we are a year later again, and Nvidia is introducing the successor to the RTX 3080 Ti, but also lower models. On this day, five GPUs intended for mobile computers officially go on the market. As the figure shows, it starts from the RTX 4090 as the most powerful model and ends with the RTX 4050 at the tail of the performance offer.
The position of the GeForce RTX 3080 Ti is taken over by the RTX 4080. It has the same number of CUDA cores (7424), but we have a narrower memory bus (192 bits versus 256 bits) and there is also a smaller memory capacity of 12 GB versus 16 GB for the older version. In terms of performance, it will probably be very similar, but this does not apply to games with DLSS 3. We will get to that later.
The flagship is of course the RTX 4090, its specifications are truly monstrous. It offers 9728 CUDA cores, 16 GB of GDDR6 memory and a frequency of up to 2 GHz. Compared to the mobile RTX 4080, the increase in computing cores is more than 30 percent, which could theoretically mean significantly higher performance. Of course, we know that the limit for similarly powerful GPUs is the processor, only tests will show the reality.
