NVIDIA’s GeForce RTX 3080 Ti is expected to be announced on 1 June, thanks to the brand’s recent devoid-of-detail teaser clip posted on Twitter. On that note, it appears that the card’s details have more or less been confirmed on GeekBench.
According to the site’s product page, the RTX 3080 Ti comes with 12GB of GDDR6X graphics memory running at 9.5GHz or 19Gbps, while its boost clock is being listed at 1.67GHz, which is slightly lower than the 1.69GHz that was initially reported. Moreover, the card is expected to have a bigger 384-bit memory bus, versus the 320-bit memory bus that the current RTX 3080 uses.
Moving on, the RTX 3080 Ti is also listed as having 80 Multiprocessors, meaning that the card also has 10240 CUDA cores. On that note, the CUDA scores of the card is listed as 238603, which is very similar to that of the RTX 3090. For context, the RTX 3090 scored 237972, which is about 0.2% greater than the RTX 3080 Ti’s score.
Interestingly enough, the testbench used to conduct and obtain the benchmark was an Intel-based system. More specifically, an 11th generation Core i9-11900K CPU, 32GB dual-channel DDR4 RAM, an ASUS ROG Maximus XIII Hero motherboard.
While the RTX 3080 Ti is expected to launch alongside the RTX 3070 Ti next week, in-depth reviews are rumoured to follow just a few days after. So far, we’ve heard nothing from NVIDIA regarding the product, but we’re definitely hoping we’ll get to review it soon.