Edge & 5G
Engage with our experts on Edge & 5G topics
Announcements
The Intel sign-in experience is changing in February to support enhanced security controls. If you sign in, click here for more information.
72 Discussions

Intel’s AI Solution Validation Lab Helps ISVs & OEMs Optimize Solutions & Accelerate Time to Market

David_Raske
Employee
5 1 1,267

Intel’s AI Solution Validation Lab Helps ISVs & OEMs Optimize Solutions & Accelerate Time to Market

David Raske is a Segment Development Manager for Smart Cities (Safety and Security) in Intel’s Network and Edge Solutions Group. In this blog, he shares details on a new initiative between Intel and Dell Technologies to accelerate AI solution validation.

 

Highlights

  • To accelerate time to market for new AI solutions, Intel and Dell Technologies created an AI solution validation lab for ISVs to optimize on the latest Intel servers.
  • EPIC iO, WaitTime and Irisity have leveraged the initiative thus far to help speed validation in the solution development cycle.
  • Reach out to your Intel representative today to get started.

 

AI Solution Development Challenges for ISVs and OEMs  

As AI and computer vision increasingly play a mission-critical role in virtually every industry, business success depends on use case-specific solutions, accurate performance estimates for de-risked deployments, and total cost of ownership (TCO) leadership.  The pace of innovation is accelerating, from AI developers creating new models and integrated software, to faster microprocessors, software tools, and on-chip features that enable even more performant processing of AI pipelines.

Across many industries, including smart cities, public safety and security, transportation, healthcare, retail, and manufacturing, technologists continually identify new use cases for AI and intelligent video that may require updated solutions for each specific combination of applications and workload.  To serve this emerging need, Original Equipment Manufacturers (OEMs) and Independent Software Vendors (ISVs) work together to combine hardware and software to create market-ready AI solutions, and then validate these to ensure stability for easy deployment.

However, with this rapid pace of change, it can be difficult and costly for the AI ecosystem to stay on the cutting edge.

 

Validation and Optimization Lab

To address this challenge, Intel and Dell Technologies co-created an initiative to help accelerate the time to market for ISVs that are ready to transition to next-gen Intel® servers.  Once ISV participants complete onboarding, ISV engineering teams are provided complete remote access to Dell edge servers hosted in an Intel lab, with Intel technical resources supporting their validation and optimization work in real time.

After the ISV validates their software on the system (in addition to making BIOS or other server changes), the optimization work begins.  While testing real video analytics workloads in the lab, ISV engineers work closely with Intel architects and OpenVINOTM toolkit experts to identify potential performance improvements. These optimization opportunities vary by each ISV but typically include further leveraging Intel AI inferencing features like Deep Learning Boost (VNNI) and AVX-512, as well as OpenVINOTM optimizations.

When testing is complete, Intel co-creates a sizing guideline report which details the system specifications as well as the final optimized performance, in terms of number of video streams supported by the particular server configuration (at various video resolutions and frames per second). With this sizing information in hand, sales and integration channel partners can confidently design high-performance and cost-effective AI video solutions around a predetermined bill of materials that are ready to deploy with reduced risk.

 

Value creation for Intel partners

 

dell.png

Wayne Arvidson, Global Director of Market Development & Strategy for Dell Technologies reflected on the customer value created through this program with Intel: “We start by talking with our customers understanding what outcomes they are trying to deliver. The ‘validated design solution’ is a process that both Dell and Intel have invested in, and that is all about minimizing the risk inherent with delivering computer vision solutions that provide better quality insights, faster. This ultimately enhances our customers' ability to achieve personnel and facility safety, improve customer and employee experience, operational efficiency, sustainability, and increase revenues. And this process is an area that differentiates Dell Technologies and Intel from other vendors in the market.”

epicio.png

One of the first ISVs to complete the program was EPIC iO, a leading developer of intelligent video analytics for public sector, smart spaces and communities, smart retail, manufacturing, and public transportation. Some of the use cases they address are site security, asset monitoring, license plate recognition (LPR), supply chain optimization, and intelligent transportation.

EPIC iO is a long-time Intel® Partner Alliance member and Dell partner, and was excited about the potential performance gains and TCO reduction from Intel’s 3rd generation Xeon platform.  EPIC iO Chief Technology Officer Michael Knight said: “EPIC iO has partnered with Intel extensively to ensure our cloud-native DeepInsights AI Platform can be deployed and operated at any edge, with complete automation, on Intel® Core™ to the latest Intel® Xeon® CPUs, instantiated on Cloud, Virtual Machine (VM) or Bare Metal, without refactoring. Our combined ability to scale workloads with optimized core and thread usage provides our customers with performance they would generally assume is only available with discreet GPU.“

 

WT-Logo-1080_250.png

Also in the lab was crowd intelligence leader WaitTime, which helps operators understand crowd conditions (queue wait times and crowd densities) to improve business intelligence, guest experiences, and safer spaces.  They serve the sports & entertainment, retail, and transportation industries, helping venues like stadiums, amusement parks, convention centers, large malls, and airports.

Dr. Thomas Sterling, AI Lead for WaitTime noted: “WaitTime leverages oneAPI and VTune™ to optimize our patented artificial intelligence. The granular documentation on code hotspots not only allows us to maximize code performance, but also allows us to optimize processing on Intel hardware.“

 

irisity-horiz.jpg

For almost 20 years, Irisity has been at the forefront of innovation in the video analytics marketplace, delivering high quality, cutting-edge video analytics products and serving diverse markets such as municipalities, transportation, critical infrastructure, and central monitoring stations.  About Irisity’s experience in the lab, Zvika Ashani, Chief Technology Officer of Irisity commented: “Irisity was able to use Intel’s  AI validation lab to quickly test and benchmark our latest IRIS+ software release on servers running a variety of 3rd Generation Intel® Xeon® Scalable Processors and create detailed hardware sizing data for our partners and customers.

 

Customer and Ecosystem Benefits

Intel silicon and software technologies enable the ISV and OEM ecosystem to develop high-performance, cost-effective, and scalable AI edge solutions. Integrators and end users benefit from these pre-validated solutions deployed on the latest Intel and OEM platforms through stable, predictable performance on pre-tested workloads, which helps reduce TCO due to optimizations that leverage advanced capabilities from Intel.

 

Future Plans

In the near future Intel plans to extend the initiative to other strategically-aligned OEMs and ISVs.  Please reach out to your Intel representative with any questions around this lab initiative or solutions mentioned here.

Learn more about intelligent video at Intel:  https://www.intel.com/content/www/us/en/internet-of-things/computer-vision/intelligent-video/overvie...

 

 

More about OpenVINO™

The Intel® Distribution of OpenVINO™ toolkit1 is a tool suite for high-performance, deep learning aimed at delivering faster, more accurate results deployed into production across Intel® architecture from the edge to cloud.  The toolkit enables a write-once, deploy-anywhere approach to deep learning deployments on ALL Intel platforms. OpenVINOTM allows developers to convert and optimize models, deploy across a mix Intel hardware and environments, on-premise and on-device, in the browser or in the cloud.

1Intel and the Intel logo are trademarks of Intel Corporation or its subsidiaries.

 

Notices and Disclaimers

Intel® technologies may require enabled hardware, software or service activation.

No product or component can be absolutely secure.

Your costs and results may vary.

Intel does not control or audit third-party data. You should consult other sources to evaluate accuracy.

Intel is committed to respecting human rights and avoiding complicity in human rights abuses. See Intel’s Global Human Rights Principles. Intel’s products and software are intended only to be used in applications that do not cause or contribute to a violation of an internationally recognized human right.

1 Comment
FrankMcCarthy
Employee

This is a great example of Intel working with our partners to drive outcome based results for our partner ecosystem! Nice job David!

About the Author
David Raske is Segment Development Manager for Smart Cities (Safety and Security) within Intel’s Network and Edge Solutions Group. His past experience includes 15 years’ security industry leadership in marketing, business development, and product roles, with a focus on IP Video and AI/ML video analytics solutions. Prior to joining Intel, David was Chief Marketing and Business Development Officer and founding member of Bastion Security, a remote video monitoring managed services provider. Earlier in his career he worked in various technical roles, including IT management in the NBA and software development at Cisco Systems. David holds an M.B.A. from The Wharton School of the University of Pennsylvania and a B.S. in Management Information Systems from Santa Clara University.
Authors