Home / Tech News / Featured Tech News / A new SKU between the RTX 3080 and 3090 is allegedly in the works

A new SKU between the RTX 3080 and 3090 is allegedly in the works

It looks like the presumed “RTX 3070Ti” is not the only new SKU that Nvidia is working on. It seems a new SKU between the RTX 3080 and the flagship RTX 3090 is also reportedly in the works.

As per @kopite7kimi, the new GA102 (codename GA102-250) card features 9984 CUDA cores, meaning it should come with 78 streaming multiprocessors (SMs) with 78 RT cores, 312 TMUs, and 312 Tensor cores. Memory wise, it’s expected to feature a 384-bit memory bus and GDDR6X modules. This leads us to believe that it should come equipped with 12GB or 24GB of GDDR6X.

If Nvidia chooses to equip it with 24GB of GDDR6X, this SKU would surely be too similar to the RTX 3090, with the only relevant difference being the inferior number of CUDA cores (512 less). With 12GB and 9984 CUDA cores, or 2GB and 1280 CUDA cores more than the RTX 3080, this supposed “RTX 3080Ti” would differentiate itself enough from the rest of the RTX 30 series line-up.

As we’ve said before, Nvidia might be rearranging their graphics cards line-up to better compete with the upcoming Radeon RX 6000 series. If the “RTX 3070Ti” and the “RTX 3080Ti” actually launch, Nvidia would have four graphics cards based on the GA102 GPU, without taking into account any possible upcoming mobile graphics cards. With the stock of the RTX 3080 and RTX 3090 being as bad as it is, manufacturing two more graphics card using the exact same GPU might strain the production lines even further.

KitGuru says: How does this “RTX 3080Ti” look to you? Would it be worth considering for those looking for a new graphics card?

Become a Patron!

Check Also

Luke & Leo (Ep9) – Intel attack AMD in media presentation

Leo recently sat through an Intel briefing relating to benchmarking laptops and focusing on performance both with an without the power adapter connected. Throughout, Intel was keen to highlight the AMD drops performance by a higher percentage in their own testing. Luke and Leo discuss this briefing and wonder why Intel is so keen to publicise AMD's weaknesses rather than focusing on Intel's own strengths.