SILICON VALLEY, CA and PARIS, FRANCE and SEOUL, SOUTH KOREA--(Marketwired - August 02, 2016) - Orange Silicon Valley, a Silicon Valley Business Innovation Center for global telecom operator Orange, and CocoLink Corp, a spinoff of Seoul National University, have built a functional prototype of one of the world's highest density Deep Learning Supercomputer in a box using CoCoLink's KLIMAX 210, a server designed for Exascale. They were able to load 20 Functional GPUs in a single 4 Rack Unit Sized server. With 20 NVIDIA K40 GPUs set at overclock (GPU boost 2) mode, the system is capable of delivering a screaming 100 TeraFLOPS in a single box with 57,600 cores. With specially engineered high performance heat sinks, this pushes the limit of computational density in any server without resorting to liquid cooling.

The A.I. researchers of Orange in France were also able to use Caffe, the popular deep learning framework to test the system for scalability. They were able to scale the training job to 16 GPUs. This endeavor is continuing with various partners to adapt the framework to its full potential to exploit all the 20 GPUs in the system. The next step would be to scale to a cluster.

The team (Orange Silicon Valley and CoCoLink Korea) has also upgraded the system with the latest commercially available NVIDIA GPUs -- GeForce GTX 1080 based on Pascal architecture. They were the first to validate a GTX 1080 for Deep Learning and identified that these consumer grade GPUs capable of achieving the same task of running GoogleNet on Caffe with 3.5 times faster speed in reaching a certain level of accuracy of image recognition during training than the NVIDIA Tesla K40 enterprise grade GPUs, which were unveiled in 2014.

This gives us a sense of how efficiency in deep learning systems is increasing over years in a beyond linear fashion.

Having identified this disruptive price/performance value proposition, the team loaded the KLIMAX system with 10 GTX 1080 GPUs.

They were able to fire up all Pascal GPUs on overclock (Boost) mode with a theoretical aggregate computation capability of 106 TeraFLOPS (Single Precision). So far the A.I. research team of Orange France were able to scale Caffe (NVIDIA fork) to 8 GPUs with beta release of CUDA 8.0 and CuDNN 5 and CuDNN4. The eventual objective is to scale the server capability with 20 Pascal GPUs with a computational horsepower in the excess of 200+ TeraFLOPS -- a feat that has never been accomplished before with consumer grade graphics card.

A particular training job on ImageNet data which used to take Orange researchers one and a half days (36 hours) with a single NVIDIA K40 can now be accomplished in 3.5 hours using 8 NVIDIA GTX 1080 cards. This is more than 10X increase in speed in regard to training performance.

As the world transitions towards Exascale and A.I. turns out to be a global race, this particular experiment is a partnership between researchers of 3 countries -- USA, France and South Korea -- working together to accelerate Artificial Intelligence by building a supercomputer in a single server by pushing the limits of thermodynamics, geometry and price vs. performance efficiency.

This currently remains as a research project for Orange and there are no plans at present to implement or develop this as a commercial offering. Detailed benchmark data based on this research will be published by the team in the near future as they make more progress towards optimization of the Deep Learning framework in collaboration with the open source community, academia and industry partners.

About Orange

Orange is one of the world's leading telecommunications operators with sales of 40 billion euros in 2015 and 155,000 employees worldwide at 31 March 2016, including 96,000 employees in France. Present in 28 countries, the Group has a total customer base of 252 million customers worldwide at 31 March 2016, including 191 million mobile customers and 18 million fixed broadband customers. Orange is also a leading provider of global IT and telecommunication services to multinational companies, under the brand Orange Business Services. In March 2015, the Group presented its new strategic plan "Essentials2020" which places customer experience at the heart of its strategy with the aim of allowing them to benefit fully from the digital universe and the power of its new generation networks.

Orange is listed on Euronext Paris (symbol ORA) and on the New York Stock Exchange (symbol ORAN). For more information on the internet and on your mobile:,, or to follow us on Twitter: @orangegrouppr and @orange.

Orange and any other Orange product or service names included in this material are trademarks of Orange or Orange Brand Services Limited.

About CoCoLink

CoCoLink, a subsidiary of Seoul National University, was established in 2001 to be an industry-leader in high-performance computing (HPC). CoCoLink continues to innovate extreme compute density to the HPC and deep learning markets by pushing the limits of server design. With innovative approaches to system integration, application development, interconnects, and processor design, CoCoLink is dedicated to push development that will one day enable future exascale computing. CoCoLink is a pioneer of GPU Centric Computing with technology that enables up to 20 GPUs in a single node.

For more information on the internet and on your mobile:, or to follow us on Twitter: @cocolink.

CoCoLink and any other CoCoLink product or service names included in this material are trademarks of CoCoLink.

For more information or to schedule an interview with any of the respective representatives, please contact and

Image Available:
Image Available:
Image Available:
Image Available:
Image Available:

Contact Information:


Dongjo Ha
Global Business Manager
+1(408) 394-3157