Last month, our article about NVIDIA’s ground-breaking new Volta architecture hinted that there may also be a new GeForce graphics card in the works, the GTX 1070 Ti. We can confirm that this new GPU was in fact officially launched in the latter half of October after much deliberation over its existence. We’re really excited by this news, considering the demand we’ve been seeing for the GTX 1080 Ti GPUs within our products over the past few months.
While as yet we don’t have the full details of NVIDIA’s latest creation to share following the launch, what we can confirm is that the GPU will sit comfortably in-between their high-end GTX 1080 and GTX 1070 cards. Its performance, though, does leave its older sibling, the 1070, in the dark – this will instead come close to rivalling the 1080 at only around 5% short of its power.
Its memory capability won’t quite reach the dizzying heights of the 1080 and 1080 Ti, however, with 8GB of GDDR5 rather than GDDR5X memory, this time resembling the 1070’s specs.
The remainder of its specs continue to borrow from its close relatives:
- Pascal architecture
- Base clock of 1607 MHz
- Boost clock of 1683 MHz
- 8GB GDDR5 memory
- 180 TDP
- 2432 CUDA cores
- Only one streaming multiprocessor (SM) disabled
Leading up to the new GPU’s launch, a lot of speculation surrounded whether NVIDIA is blocking overclocking for the GTX 1070 Ti – as far as we know, they will not be. It’s possible that they will be locked for hardware manufacturers on release only, but our research indicates that manual overclocking will still be enabled.
We’re keeping our ears to the ground for any further news on the GTX 1070 Ti, but in the meantime we’ll be looking to see how we can best accommodate the new GPU within our products in order to optimise its graphics potential for our customers.