Hardware company Nvidia has announced a new GP102-based Titan X graphics card that is up to 60 percent faster than the previous Titan X model.
The GPU was revealed at a special artificial intelligence presentation
at Stanford University, with Nvidia CEO Jen-Hsun Huang making the
announcement.
Specifications are as follows:
12-billion transistors
11 TFLOPs FP32 (32-bit floating point)
44 TOPS INT8
3,584 CUDA cores at 1.53GHz (versus 3,072 CUDA cores at 1.08GHz in previous TITAN X)
High performance engineering for maximum overclocking
12GB of GDDR5X memory (480GB/s)
This
new Titan X is based on Nvidia's new Pascal architecture, and houses
3,584 CUDA cores--the highest amount ever. The card will be available
from the Nvidia website
for North America, Europe, and "select system builders" from August 2.
It will be "coming soon" to Asia. If you're thinking of picking one up,
it'll set you back $1,200. Note that Nvidia has also called it "the
biggest GPU ever built" so it might be worth checking if it'll fit your
build before making a purchase.
GameSpot recently reviewed Nvidia's GeForce GTX 1060, GTX 1070, and GTX 1080 line of cards. Nvidia first revealed the GTX 1060 in July, a $250 card that the company claimed to deliver "GTX 980-level performance."
