Ok so here is a quick one. What does “Ti” mean when seen in a NVIDIA GPU Name?
Well its actually the periodic table abbreviation for “Titanium”.
It doesn’t mean however, that there is any Titanium used in the fabrication of the GPU. At least not more than in any non-Ti cards.
Now how does this translate to rendering performance? Let’s take a look at some of the more recent generations of NVIDIA Cards:
- GTX 780: CUDA-Cores: 2304, VRAM: 3GB, Octane-Bench: 73
GTX 780Ti: CUDA-Cores: 2880, VRAM: 3GB, Octane-Bench: 91
- GTX 980: CUDA-Cores: 2048, VRAM: 4GB, Octane-Bench: 98
GTX 980Ti: CUDA-Cores: 2816, VRAM: 6GB, Octane-Bench: 134
- GTX 1080: CUDA-Cores 2560, VRAM 8GB, Octane-Bench: 136
GTX 1080Ti: CUDA-Cores: 3584, VRAM 11GB, Octane-Bench: 185
As you can clearly see, the Ti Versions have more Shader Cores, usually a higher Clock Speed, more VRAM and tend to be the best performance / price ratio and performance / power ratio GPUs in the NVIDIA GPU Generation line-up.
The Ti Version also clearly outranks the non-Ti GPU in Octane-Bench.
Some say the Ti Version usually is a non-cut down Version of a GPU Core. Though this is not always true. Sometimes the Ti Version uses a different Core than the non-Ti Versions, though sometimes the non-Ti Versions have the Same GPU Core as the Ti Version with some parts being deactivated.
Check the performance / Dollar Table here to see if you should reach for the Ti Version of NVIDIA GPUs.