While NVIDIA is counting down to its next generation GeForce GPU launch at Gamescom 2018, images of the alleged GeForce RTX 2080 and RTX 2080 Ti from the AIB partners has already been spreading on the internet like wildfire.
Instead of staying with the usual GTX suffix, NVIDIA now uses the new RTX suffix for its new GeForce GPU. The teaser was published on NVIDIA’s official Youtube channel on August 13th, with quite an amount of hints:
- RoyTeX (the RTX suffix)
- not_11 (there will be no GTX 1180)
- gimme 20 and Mac-20 (it’ll be RTX 2080)
- August 20th (the launch date)
Leaked Photos: GeForce RTX 2080 and RTX 2080 Ti
The first leaked image of the alleged GeForce RTX 2080 and RTX 2080 Ti from MSI including the box that has the suffix (RTX) and the model (2080 and 2080 Ti) was published on videocardz.com, and you’ll can probably expect to see more later on.
Other leaked image of the GeForce RTX 2080 and RTX 2080 Ti from other NVIDIA AIB partners such as ASUS, EVGA, PNY, ZOTAC, Gigabyte, Palit and Inno3D was subsequently added to the list, alongside more speculated specifications of both the GeForce RTX 2080 and RTX 2080 Ti.
|GPU Model||12nm TU102|
|Cores : TMUs : ROPs||4352 : 184 : –|
|Effective Memory Clock||14000 MHz|
|Computing Power (FP32)||15,232 GFLOPS|
|Memory Size||11264 MB GDDR6|
|Memory Bus Width||352-bit|
|Memory Bandwidth||616 GB/s|
|Interface||PCI-Express 3.0 x16|
|Thermal Design Power||250W|
According to videocardz.com, the GerForce RTX 2080 Ti will pack a 11GB GDDR6 memory clocked at 14Gbps, 352-bit memory bus width and memory bandwidth of 616 GB/s. Compared to the recently released Quadro RTX 6000 and RTX 8000, the amount of CUDA cores that RTX 2080 Ti has (4352 CUDA cores) is slightly fewer than both the Quadro (4608 CUDA cores).
As for the GerForce RTX 2080, lower CUDA core count and memory bus width is pretty much expected. Given that the speculated specifications is correct, 2944 CUDA cores, 256-bit memory bus width and 448 GB/s memory bandwidth is what you’ll get on the RTX 2080.