erek
[H]F Junkie
- Joined
- Dec 19, 2005
- Messages
- 9,788
Pretty nice
"The billion-dollar question hence would be exactly how NVIDIA segments the three SKUs—whether it stops at just memory size and memory bus width; or whether it will also tinker with core-configuration, and possibly even base one of the three SKUs on a physically smaller silicon, such as the AD106. The "RTX 4070" rumored to be a single SKU until now, was expected to be carved out of the AD104, with 5,888 out of 7,680 CUDA cores being enabled. NVIDIA faced stiff criticism from the media and gamers for such segmentation for its RTX 4080 16 GB and RTX 4080 12 GB, with the company being forced to cancel the market-release of the latter, and relaunch it under the name RTX 4070 Ti."
Source: https://www.techpowerup.com/304976/...-16gb-variants-of-rtx-4070-gigabyte-thinks-so
"The billion-dollar question hence would be exactly how NVIDIA segments the three SKUs—whether it stops at just memory size and memory bus width; or whether it will also tinker with core-configuration, and possibly even base one of the three SKUs on a physically smaller silicon, such as the AD106. The "RTX 4070" rumored to be a single SKU until now, was expected to be carved out of the AD104, with 5,888 out of 7,680 CUDA cores being enabled. NVIDIA faced stiff criticism from the media and gamers for such segmentation for its RTX 4080 16 GB and RTX 4080 12 GB, with the company being forced to cancel the market-release of the latter, and relaunch it under the name RTX 4070 Ti."

Source: https://www.techpowerup.com/304976/...-16gb-variants-of-rtx-4070-gigabyte-thinks-so