The GeForce RTX 4080 could arrive in two versions with 12 GB and 16 GB of VRAM

geforce rtx 4080 portada 1000x600.jpg
geforce rtx 4080 portada 1000x600.jpg

One of the most anticipated and most important graphics cards within the new graphics generation that will arrive at the end of this year is the GeForce-RTX 4080a model that will succeed the GeForce RTX 3080 and that according to the latest information could be available in two different configurations, one with 12 GB of graphics memory and one with 16 GB.

It would not be the first time that NVIDIA has launched a high-end graphics card with two memory configurations, in fact if we pull memory we realize that it has been a fairly common practice. Remember, for example, that the GeForce GTX 580 was available in two versions, one with 1.5 GB and one with 3 GB, and the same thing happened with the GTX 680, which could be bought in versions with 2 GB and 4 GB of graphics memory. Within the current generation we have also seen two versions of the GeForce RTX 3080, one with 10 GB and another with 12 GB of VRAM.

The GeForce RTX 4080 with 12 GB of GDDR6X will use a 10 layer PCB and the GeForce RTX 4080 16GB GDDR6X will have a 12 layer PCB. For comparison purposes, we remind you that according to the most recent rumors, the GeForce RTX 4090 will have a 14-layer PCB. We are always talking about the reference models, that is, the Founders Edition, so keep that in mind, as it is very important.

Although the rumors do not agree, everything seems to indicate that the presentation of the Ada Lovelace architecture will take place at this year’s GTC, which will be held at the end of September, and that the launch of the GeForce RTX 4080 would take place in October. Other information speaks of a delay and a possible launch between November and December, but the truth is that there is nothing definitive yet.

SEE ALSO  Can I connect to my neighbor's WiFi printer to play a prank on him?

GeForce-RTX 4080

Possible specifications of the GeForce RTX 4080

  • AD103 graphic core at 5nm node from TSMC.
  • 9,728 shaders at 2,025-2,375 MHz.
  • 304 texturing units.
  • 304 tensor cores for AI.
  • 96 raster units. There are rumors of a huge bump up to 214 raster units.
  • 76 RT cores for ray tracing.
  • 256 bit bus.
  • 12 GB or 16 GB of 23 GHz GDDR6X graphics memory (736 GB/s of bandwidth).
  • 340 watt TGP.

The latest leaks indicate that the GeForce RTX 4080 reaches a performance of 15,000 points in 3DMark Spy Extremewhich means that it would be quite a bit more powerful than the GeForce RTX 3090 Ti, which averages a score close to 11,000 points. If this is confirmed, it is clear that Ada Lovelace is going to represent a significant leap in performance against Ampere.

However, it is important to remember that numerous rumors have also taken for granted that this new generation will greatly improve ray tracing performance, so much so that the GeForce RTX 40 could end up being twice as powerful with that workload than the current generation. Sounds so good it’s hard to believe, but considering NVIDIA matured RT cores with Ampere it’s not something we can rule out either.