'%20x='0'%20y='0'%20height='100%25'%20width='100%25'%20%0A%20%20%20%20%20%20%20%20%20%20xlink%3Ahref='data:image/jpg;base64,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'%3E%3C/image%3E%3C/svg%3E)
Nvidia recently announced the RTX 4080 series of GPUs in the GTC Sept 2022 Keynote. Said to provide two to four times more performance than the last generation RTX 3080 Ti, these new graphics cards will be based on the 'Ada Lovelace' architecture, which on paper has exponential performance increments over the last generation of 'Ampere' architecture.
Based on TSMC 4N Process and Micron G6X Memory, the 3rd Gen RT cores on these cards will allow up to two times the ray tracing performance of last gen GPUs. Apart from this, the new Tensor cores will enable deep learning capabilities unlike anything seen before.
RTX 4080 GPUs will be available for purchase in two variants – 12 GB and 16 GB. The former model of the GPU will start at $899, while the latter from $1,199 in the United States.
However, the naming policy adopted by Nvidia for their RTX 4080 variants can be misleading.
Everything users need to know about the RTX 4080 variants
As previously mentioned, the RTX 4080 will be arriving in two different variants. Nvidia has had these two variants separated only on the basis of VRAM, but there's more to it than just that.
Other than the memory size of the two variants, these two GPUs differ in these aspects:
The 12 GB variant has almost 24% lesser CUDA cores compared to the 16 GB variant. Thus, this gives the 12 GB sport not only lesser video memory but lower graphic processing capabilities. As a result, both variants won't be delivering equal performance, even in non-memory bound scenarios.
Performance difference between the 12 GB and 16 GB variants
Performance numbers reported by Nvidia themselves clearly show the difference in the frames per second that both the variants are capable of producing. As seen in the graph, the 12 GB variant is only around 1.5x faster than the last gen RTX 3080 Ti in Warhammer 40,000: Darktide. However, the 16 GB variant is almost 2x faster.
Similarly, in Cyberpunk 2077, the 12 GB variant is around 2.5x faster than the last generation GPU. Meanwhile, the 16 GB variant delivers 3x the performance of RTX 3080 Ti.
However, it appears that there isn't much difference in Microsoft Flight Simulator. But again, there's a catch. Microsoft Flight Simulator is a CPU-heavy game, and as a result, users may experience CPU limitations, which could provide similar performance.
Conclusion
Nvidia has previously deployed such naming strategies with their 1000 series of cards. GTX 1060 came in 3GB and 6GB models. Although both the GPUs were named GTX 1060, the 3GB variant lacked severely in performance compared to the 6GB model. Similarly, both RTX 4080 variants come under the same name, but they aren't quite the same in terms of performance. This can mislead customers.
Consumers will only be observing the VRAM difference as that is what's being marketed. Since it isn't much, they might go for the cheaper version assuming they provide the same performance straight out of the box.
RTX 4080 GPUs will be available for purchase this November, 2022. While no specific dates have been announced, gamers are speculating a late November release.
