This year Cisco was a Platinum sponsor of the GPU Technology Conference in San Jose. Colloquially known as “GTC2016” or “The Graphics Conference,” it gave us a great platform to show off some of the latest developments from Cisco and our partners.
We showed a host of applications and solutions along with NVIDIA and our ecosystem partners:
Hugo Latapie, Cisco Principal Engineer, presented a session on Deep Learning at the Edge of the Network. He highlighted use cases in IoT, SmartCities, retail, event analytics, and transportation. His talk was targeted primarily towards engineers either actively engaged in product development or seriously contemplating it.
The idea is to use these techniques of data collection at the edge (fog computing) along with virtualized graphics from NVIDIA within the high performance Cisco UCS data center computing environment along with deep learning algorithms in visual, crowd, and behavioral analytics projects at Cisco. Why? Deep Learning is used to improve a whole bunch of applications for vehicle, pedestrian, and landmark identification for driver assistance – and for Image recognition and Life Sciences.
Cisco also works with application vendors to deliver validated solutions for deploying highly specialized 3D applications virtually. We have solutions for the PTC Creo Suite (Manufacturing) and Esri ArcGIS Pro (geospatial mapping) among others. Last year we heard how Ford is using NVIDIA GPUs in their Cisco UCS infrastructure and are improving engineering/design productivity by partnering with Citrix.
This year there was a new message from Ford. Last year they shared it can take 7 hours for a designer to download the data sets he or she needs to work on to do ten minutes work. Now they are saying:
“You don’t need to move your data to an expensive workstation possibly out-of-your control or out of the country to get work done”
This year we hosted our partner Citrix in our booth to show how professionals can avoid expensive workstations with the use of technology.. Citrix demonstrated the latest remote graphics solution powered by Cisco and NVIDIA. Visitors experienced the HDX 3D Pro user experience (built on the flexible Cisco UCS platform), Citrix HDX Framehawk with GPU support, and HDX USB device redirection to support additional USB drawing tools used by professionals.
There were all sorts of new NVIDIA GRID 2.0 products on display. One standout was the new TESLA M6 MXM GPU, which brings graphics acceleration to our VDI workhorse – the UCS B200 M4 blade server. This provides UCS customers a wide range of configurations to support any VDI deployment needs they may have.
We also showed our new Cisco HyperFlex hyper-converged infrastructure (HCI) solution in the booth. Cisco HyperFlex is the first end-to-end HCI solution that combines compute, networking, and next generation storage.
We also had ESRI in our booth showing ArcGIS Pro, which shows geospatial data on desktops and is used in the utility and oil and gas industries for visualizing, analyzing, and sharing data. At last year’s Cisco Live I showed this on BitStew demo screens. The pièce de résistance here is that you can run the app miles away on a Cisco UCS datacenter, without the need for big local workstations.
Around the Event
Other folks were showing off cool technologies, and I asked our own Philip Laidlaw what he saw there this month:
“Autonomous cars (of course) and a very cool wave powered floating drone from Liquid Robotics that can head to a point anywhere in the open oceans to do data collection, and investigate with missions specific payloads.”
This would be great for the oil and gas industry.
Were you there? If so, what did you think? Share your thoughts in the comments section below.
Also, see how we can help your industry here.