The Edge: How AI Learning and Monitoring Video Data Has Matured

Data has become the common denominator. Today, data has transformed video monitoring in ways we never even imagined a decade ago.

As this data is captured, processed, and used from the nearest and farthest edges of the network, businesses like buildings, Governments, retail and more are harnessing the power of the network edge.  

Take a large commercial organization with manufacturing environment, there might be 200 CCTV cameras located along the line. Lots of data points are being gathered from each camera, for example safety vests being worn, hard hats, location of people and more all capturing and sending a multitude of data points per second.

Using AI to train at the edge is a critical part, especially at scale. Processing all incoming data and storing the information gathered can quickly get expensive. However, with specifically designed solutions that capture this data, significant savings can be realized. With tools that allow for 90%+ bandwidth cost savings, VOD to cloud efficiency, on prem edge AI processing and ultra low latency, we can monitor controls, learn via AI and better manage all connected devices within the facility. Simply beautiful advancements.

It’s time to start the conversation. What can the edge and AI video learning do to improve your organization’s operations?

The New Smart Building: Artificial Intelligence to Automate Access Control, Parking and More

AI powered solutions now can help to redefine business operations providing real-time data updates that continually learn and refine that data. The products literally get better and more refined as time progresses, something any procurement manager must love!

The Gartner Top Strategic Technology Trends for 2021 notes, anything that can be automated in an organisation should be automated. Businesses will aim to automate as many business and IT processes as possible, relying on Artificial Intelligence.

AIoT (Artificial Intelligent of Things) is at the heart of it all and one market that is exploding is smart buildings, projected to grow from USD $66.3 billion in 2020 to USD $108.9 billion by 2025.

Whether it be energy efficiency systems or building access, integration to building management platforms for an end-to-end solution is now available to also include connecting and managing all cameras to real-time edge on-prem video AI analytics.

Building operations managers now can gain deeper understanding and control of building access by each door, can manage property intrusion alerts, understand anomaly object detection, see building parking violation, even employee & customer happiness index through face recognition.

It’s time to start the conversation. What can AIoT and a new approach to smart buildings do to improve your organization’s operations?

Composable Data & Video Cameras: How the Future of Adaptive Computing is Now

Video cameras are everywhere and there will be more every day as we move forward in places like retail, buildings, cities, parking, oil & gas and many more areas of business.

The pandemic has changed so much, including forcing data centers to scale up their infrastructure to meet the performance and latency requirements of a variety of demanding workloads. The good news in all of this change, the time has never been better for the camera/CCTV world, as this much improved  AI video based evolution on composable computing platforms will now make it easer than ever to implement AI based camera solutions like smart building, smart retail or smart parking.

With an accelerated road to digital transformation and responsible, scalable AI, the future has arrived. It is no longer a pipe dream. The new reality we live in, now offers change and trends coming at us like never before. Research shows that 72% of data and analytics leaders are leading, or are heavily involved in their organizations’ digital transformation efforts. These data leaders now confront emerging trends on various fronts, every day.

As an example, managers within City and parking operations can now take advantage of much automated video data extraction and analytic technology to operate more efficiently with AI solutions that identify and learn license plate recognition and can be combined with automated payments, parking violation detection, road marking detection, and more.

It’s time to start the conversation. What can composable data & video cameras do to improve your organization’s operations?

Aupera Technologies Disrupts Online Video Processing with New Platform Powered by Xilinx

October 15, 2018 08:00 AM Eastern Daylight Time

BEIJING–(BUSINESS WIRE)–(Xilinx Developer Forum)– Aupera Technologies, Inc. a leading technology innovator dedicated in providing Video Processing Solutions in Data Centers, announced the General Availability of its Aup2600 series system based on the latest Xilinx Zynq® UltraScale+™ MPSoC, tackling the challenges faced by video streaming applications at scale requiring real-time video transcoding and video content analytics.

The AuperaTM Aup260x distributed video processing system, is a disruptive technology for data centers, as it negates the need for traditional x86 processors. A fully loaded system, contains 48 Zynq® UltraScale+™ MPSoC, and handles 380 concurrent video streams, delivering 3.3x more performance at 1/10th the energy cost, while using 1/10th the rack space compared to traditional x86 solutions.

Aupera is also closely collaborating with Xilinx to develop various Machine Learning algorithms, enabling real-time AI-based video analytics and search functionality. These new capabilities will be hardware-accelerated on the Aup260x without the need to replace or upgrade any hardware.

“Our high-density FPGA-based video processing system is now qualified for large scale deployment. It is providing a truly disruptive online video processing solution that will help our customers achieve 33x performance per watt, at a fraction of TCO, while delivering high quality video service to end users. We believe we are changing the landscape of video processing in data centers,” said Roy Liao, CEO of Aupera Technologies.

“We are extremely excited to see such innovation coming from our partner Aupera, who are using our technology and delivering a game-changing platform for large-scale video streaming applications. This is also helping us deliver on our ‘data center first’ strategy,” Said Ramine, VP of software product planning at Xilinx, “We will continue to closely collaborate with Aupera throughout the deployment, to extend the capabilities of this platform, using our adaptable hardware and AI technology.”

Aupera Technologies is showcasing their system at the XDF Beijing show, and will be giving a breakout session titled “AI power FPGA based video transcoding for real time application”.

About Aupera Technologies Inc

Aupera Technologies is a technology innovator which builds converged architecture of video transcoding, content classification, and video storage system. With our proprietary distributed computing and storage architecture, coupling with AI acceleration engine, Aupera is building next generation video data center platform.

Contacts

Aupera Technologies Inc., Yinghua Chen +1-604 563-1862 Email: info@auperatech.com

Aupera Technologies Reveals Next-Gen NVMeoF JBOF Prototype Achieving 16M IOPS

August 07, 2018 08:00 AM Eastern Daylight Time

SANTA CLARA, Calif.–(BUSINESS WIRE)–Aupera Technologies today announced an ASIC based 25G/100G Ethernet NVMeoF JBOF[3] Prototype System, which utilizes a hardware NVMeoF bridge and 25G/100G Ethernet switch fabrics, achieving native NVMe (PCIe3.0) SSD performance. Designed for scale-out, the new JBOF system significantly simplifies the deployment of high-performance storage systems in data centers.

This JBOF system is a standard 19-inch 2U chassis and supports up to 24 NVMe-oF SSD modules. The NVMe-oF SSD module adopted the interposer solution, jointly developed by Aupera and Toshiba Memory Corporation. The interposer is based on the Marvell® 88SN2400 NVMe-oF SSD converter controller, enabling dual 25Gbit RDMA network ports and NVMe-oF bridging. The built-in dual-star switch matrix provides dual 6 x 100G uplink ports and dual 24 x 25G downlink ports to SSDs for non-blocking access. The prototype system achieves 16 million IOPS[3] (4KB random read), while maintaining extremely low power consumption. The JBOF chassis has also integrated the BMC system for chassis configuration, health monitoring and remote management, which is ready for carrier grade deployment.

“With this whole new system architecture, combined with Aupera’s existing high density FPGA-based video processing solutions, we have achieved another mile-stone step towards building a next-generation video data center platform. It has been a great pleasure to work with Toshiba and Marvell on the most advanced technologies in the industry. We are very excited about the system performance achieved so far,” said Aupera CEO, Roy Liao.

NVMe over Fabrics is expected to be widely adopted in the cloud data center and enterprise markets. Toshiba Memory’s near-native performance of NVMe SSDs, combined with Aupera’s NVMeoF JBOF system, will ensure SSDs peak performance, simplify deployment, and facilitate large-scale adoption in the data center.

Aupera Technologies and Toshiba Memory will demonstrate this new JBOF prototype system at the Flash Memory Summit in Santa Clara, California from August 7 – 9 in booth #951(Aupera) and #307(Toshiba).

*All product names and service names mentioned herein may be registered trademarks or trademarks of their respective companies.

About Aupera Technologies Inc

Aupera Technologies is a technology innovator which builds converged architecture of video transcoding, content classification, and video storage system. With our proprietary distributed computing and storage architecture, coupling with AI acceleration engine, Aupera is building next generation video data center platform.

Notes:

[1] NVMe™ is a trademark of NVM Express. PCIe® is a registered trademark of PCI-SIG. All other company names, product names, and service names mentioned herein may be either registered trademarks or trademarks of their respective companies.

[2] IOPS: Input Output Per Second (or the number of I/O operations per second)

[3] JBOF: Just A Bunch of Flash

*Information in this document, including product prices and specifications, content of services and contact information, is correct on the date of the announcement but is subject to change without prior notice.

Contact

Aupera Technologies Inc., Yinghua Chen, +1-604-563-1862 Corporate Communications yinghua@auperatech.com

Social media & sharing icons powered by UltimatelySocial