Tag: deep-learning

Cray’s new XC50 supercomputer hits one petaflop of computing in a single cabinet

 Cray has a new supercomputer called the XC50, the successor to its XC40 model and the first supercomputer from the company that can deliver one petaflop of performance (at peak) in a single cabinet. The XC50 supports the NVDIA Tesla P100 GPU accelerator, as well as nex-gen Intel Xeon and Xeon Phi processors, and is targeted at uses including deep learning algorithms, which Cray says are… Read More

Continue reading here:
Cray’s new XC50 supercomputer hits one petaflop of computing in a single cabinet

NVIDIA’s made-for-autonomous-cars CPU is freaking powerful

NVIDIA debuted its Drive PX2 in-car supercomputer at CES in January, and now the company is showing off the Parker system on a chip powering it. The 256-core processor boasts up to 1.5 teraflops of juice for “deep learning-based self-driving AI cockpit systems, ” according to a post on NVIDIA’s blog . That’s in addition to 24 trillion deep learning operations per second it can churn out, too. For a perhaps more familiar touchpoint, NVIDIA says that Parker can also decode and encode 4K video streams running at 60FPS — no easy feat on its own. However, Parker is significantly less beefy than NVIDIA’s other deep learning initiative, the DGX-1 for Elon Musk’s OpenAI, which can hit 170 teraflops of performance. This platform still sounds more than capable of running high-end digital dashboards and keeping your future autonomous car shiny side up without a problem, regardless. On that front, NVIDIA says that in addition to the previously-announced partnership with Volvo (which puts Drive PX2 into the CX90), there are currently “80 carmakers, tier 1 suppliers and university research centers” using Drive PX2 at the moment. For the rest of the nitty-gritty details, be sure to hit the source link below. Source: NVIDIA

Excerpt from:
NVIDIA’s made-for-autonomous-cars CPU is freaking powerful

NVIDIA’s new top-end graphics card is the $1,200 Titan X

If you recently bought a $599 NVIDIA GTX 1080 in order to have the fastest rig around, I have bad news. NVIDIA has revealed the latest Titan X , a graphics card with 12GB of GDDR5X memory and 3, 584 cores running at 1.53 GHZ, yielding an absurd 11 teraflops of performance. That easily bests the 8.9 teraflops of the GTX 1080, which itself put the last-gen Titan X to shame . You probably won’t feel too bad, however, when we tell you that the new card has a price tag of $1, 200, double that of its now-second-best sibling. The Titan X has 12 billion transistors and runs at 250W, meaning it burns around 40 percent more power than the GTX 1080. It carries DisplayPort 1.4, HDMI 2.0b and DL-DVI ports, though the company hasn’t yet detailed the configuration. NVIDIA has now unveiled four cards (the GTX 1060 , 1070 , 1080 and Titan X) in just over two months, which is a pretty frenetic launch rate. To hammer home the point about brute horsepower, NDIVIDA CEO Jen-Hsun Huang did a surprise unveil of the Titan X at a meetup of artificial intelligence experts at Stanford University. That’s fitting, because it’s starting to blur the line between its gaming cards and Tesla GPU accelerators used for deep learning in servers and supercomputers. The card will go on sale August 2nd in North America and Europe for $1, 200, but only on NVIDIA’s site and via “select system builders.” Source: NVIDIA

View post:
NVIDIA’s new top-end graphics card is the $1,200 Titan X

The Latest Google Algorithm Creates Video Based On a Few Still Images

Google’s engineers can do some pretty incredible things with the consumer technology it has developed—from “dreaming” neural networks based on computer vision to an algorithm that can create video from Street View images. Read more…

See the article here:
The Latest Google Algorithm Creates Video Based On a Few Still Images