NVIDIA Drops Pascal Bomb With GP102-Based TITAN X, 60 Percent Faster Than Its Predecessor

pascal titan x

NVIDIA has historically taken a somewhat different approach when announcing its newest Titan-branded graphics cards, that differs from the measured, methodical launches of its more mainstream consumer products.

Tonight’s announcement, however, takes the cake. We just received details from NVIDIA regarding an upcoming ultra-powerful, Pascal-based Titan X, featuring a 12 billion transistor GPU, codenamed GP102. The email’s subject began with “SURPRISE” and the body started with, “It began with a bet”.

NVIDIA Ceo Jen-Hsun Huang With Pascal-Powered GeForce GTX Titan X

NVIDIA is obviously having a little fun with this one.

At an artificial intelligence meet-up at Stanford University this evening, NVIDIA CEO Jen-Hsun Huang first announced, and then actually GAVE AWAY a few brand new, Pascal-based NVIDIA TITAN X GPUs, much to the surprise of those in attendance.

pascal titan x 2
The New Pascal-Based NVIDIA TITAN X

Apparently, Brian Kelleher, one of NVIDIA’s top hardware engineers, made a bet with NVIDIA CEO Jen-Hsun Huang, that the company could squeeze 10 teraflops of computing performance out of a single chip. Jen-Hsun thought that was not doable in this generation of product, but apparently, Brian and his team pulled it off.

The bet was for $1. Randolf and Mortimer would be proud.

The new Titan X is powered by NVIDIA’s largest GPU; the company says it’s actually the biggest GPU ever built – period. The Pascal-based GP102 features 3,584 CUDA cores, clocked at 1.53GHz (the previous-gen Titan X has 3,072 CUDA cores clocked at 1.08GHz). Over and above the core counts, however, this new Titan X also benefits from the updated Pascal architecture. In terms of its design, it’s got a very similar aesthetic to the GeForce GTX 1080, but there’s lot more going on under the hood.

The specifications NVIDIA has released thus far include:
  • 12-billion transistors
  • 11 TFLOPs FP32 (32-bit floating point)
  • 44 TOPS INT8 (new deep learning inferencing instruction)
  • 3,584 CUDA cores at 1.53GHz
  • High performance engineering for maximum overclocking
  • 12GB of GDDR5X memory (480GB/s)

And here's how the specs of the new beast map out versus previous and current generation NVIDIA high-end GPUs:

TITAN X (New) GTX 1080 GTX 980 Ti TITAN X (Old)
GTX 980
NVIDIA GPU GP102 GP104 GM200 GM200 GM204
ArchitecturePascal Pascal Maxwell Maxwell Maxwell 
Number of Cores 3584 2560 2816 3072 2048
Core Clock 1417 MHz 1607 MHz 1000 MHz 1000 MHz 1126 MHz
Memory 12GB 8GB 6GB 12GB 4GB
Memory Clock 10000 MHz 10000 MHz 7000 MHz 7000 MHz 7000 MHz
Memory Interface 384-bit G5X 256-bit G5X 384-bit 384-bit 256-bit
Memory Bandwidth 480 GB/s 320 GB/s 336 GB/s 336 GB/s 224 GB/s
TDP 250 watts 180 watts 250 watts 250 watts 165 watts
Peak Compute 11.0 TFLOPS 8.2 TFLOPS 5.63 TFLOPS 6.14 TFLOPS 4.61 TFLOPS
Transistor Count 12B 7.2B 8.0B 8.0B 5.2B
Process Tech 16nm FinFET 16nm FinFET 28nm 28nm 28nm
MSRP $1,200 $599 $649 $999 $499

Note, that this card is not powered by the same GPU and HBM2 memory featured on NVIDIA's flagship Tesla. The architecture is the same, but the GP102 is paired to GDDR5X memory.

NVIDIA also just posted a teaser video to its YouTube channel. There aren’t any more details revealed in the video, but the visuals are interesting nonetheless.

We’re told the new Titan X will be available Aug. 2 for $1,200 direct from NVIDIA.com in North America and Europe, and from select system builders. Availability in Asia will come at a later date.