Nvidia's latest addition to the family, the Quadro P6000 can handle 432GB per second of memory bandwidth. Image Source: Tecno Gaming

Anaheim, Calif. – Nvidia (NASDAQ:NVDA) continues to push the limits on graphic performance with each new Graphical Processing Unit (GPU) on the Pascal architecture. Recently announced Quadro P6000 delivers 12 teraflops of single-precision performance.

The unit was introduced in the SiraGraph trade show, the epitome of computer and mobile graphics and virtual and augmented reality. A few blocks from Disneyland, attendees got to see the world’s first 3D oil painting simulation from Adobe Research.

The claimed fastest GPU showed up just a week after the current leader, their Titan X GPU, which, according to Nvidia, was the single most powerful GPU ever build.

Now, they exceeded their own record and raised from 11 teraflops to 12. Nevertheless, the Titan is built for top-gear gaming, whereas Quadro’s series have a different output.

The specs

The P6000 is better for demanding workstations, designed for a task like editing HD video and managing special effects. As previous models, Quadro-specific graphic drivers are optimized to accelerate interactive physically-based rendering applications, including 3D videos. Such would be the case for engineer software as Solidworks and AutoCAD.

The GPU has 3,840 cores and 24GB of GDDR5X memory. It has four DisplayPort 1.4 slot, supporting resolutions up to 4K at 120Hz and 5K at 60Hz. It also offers one DVI-D port and an optional stereo port.

The Quadro family

Nvidia also revealed the Quadro P5000, a model that delivers 8.9 teraflops of peak performance, has 2560 cores, 16 GB of GDDR5X memory and 288GBps of memory bandwidth.

The current top-end Quadro is the M6000, based on Maxwell architecture. It has a peak single-precision performance of 7 teraflops and uses GDDR5 memory. This component was first launched with a tag of $5000, a clue of the possible prices of the upcoming new models.

There are other 6 different and older models in the Quadro family for PC and 8 different Quadro GPUs designed for laptops. While Quadro models could work impressively in games, its real objective is to make workflow faster for professionals who work with immersive visuals.

For now, the P6000 is the fastest GPU around, and a similar faster-than-ever chip will follow. Main competitor AMD (NASDAQ:AMD) debuted its line of more affordable Radeon Pro WX workstation-focused GPU, available for less than $1000.

Pixar Animation Studios likes Nvidia’s latest addition

Pixar Animation Studios (NASDAQ: PIXR) has already endorsed the P6000. Pixar CTO Steve May pointed out that the studio’s artist often work with 50GB or higher datasets.

“The ability to visualize scenes of this size interactively gives our artist the ability to make creative decisions more quickly,” said Steve May. “We’re looking forward to testing the limits of Pascal and expect the benefits to our workflow to be huge.”

NVidia Pascal-based graphical units boast top-of-the-line behavior thanks to its 3840 CUDA parallel processing cores. But that kind of power demands a lot of power, as much of 250 watts, PCI Express 3.0 and 16 slots on the PC’s mainboard.

Source: Tech Times