The emphasis with the five new offerings, using technology from Addfor, Everseen, cnvrg.io, SAS and NetApp, is on the solutions’ ease of deployment, removing a barrier that has stymied many enterprises from getting AI products into production.
Today, the newly renamed Lenovo Infrastructure Solutions Group (ISG) – formerly the Lenovo DCG – is announcing the expansion of their artificial intelligence [AI] portfolio with the launch of five new AI solutions, all of which come through third party partnerships, and all of which are ready to deploy for customers.
“I worry that what we are announcing may come across a little bit boring,” said Scott Tease, Vice President and General Manager of HPC and AI, Lenovo Infrastructure Solutions Group. “It is not massively unique from an IP standpoint. But we are bringing it together in a way that’s so easy for customers to consume. It’s so hard for many of them to get beyond a Proof-of-Concept state. Knowing that these solutions will work will change that.”
Tease indicated that while Lenovo does a lot of AI-related work with emerging tech companies, big cloud users, and big research users, who see AI as an extension of what they have done with HPC, these are not the target audience for the five new solutions.
“We are looking more at early adopters in the enterprises, who don’t necessarily have Data Scientists and who have limited AI skills,” he said. “This is the focus of the announcement today.”
Lenovo’s third party ecosystem of vendors is critical to being able to bring easy to consume solutions to market quickly.
“We launched our AI Innovation Centres in 2017 to do deep engagements with customers,” said Robert Daigle, WorldWide AI Business Leader at Lenovo ISG. “We still do Proof-of-Concepts and AI workshops and benchmarking at the centres, but now many customers are looking for more turnkey solutions. So we also validate ISVs at the centres, many of them startups. We vet these partners, to give customers a sense of security.”
One of the collaborations is with Turin Italy-based Addfor, and is an AI solution to track COVID-19 safety protocols. It combines Addfor’s crowdHEDGE AI solution and Lenovo ThinkSystem SE350 edge servers powered by NVIDIA T4 Tensor Core GPUs, using AI drone footage to extract data insights in outdoor spaces like city parks.
“What’s really compelling is that it can be used beyond this current use case related to the pandemic,” Daigle said. “It can also do things like PPE scanning for manufacturing, to make sure workers are properly wearing protective gear.”
One of the non-startup collaborations announced is with veteran enterprise analytics software vendor SAS.
“This is an edge AI analytics solution, and SAS is the leading company in this space,” Daigle indicated. “They have great technology in analytics and AI, and their recently announced SAS ESP platform targets AI and analytics workloads at the edge.”
This Edge AI solution combines a Lenovo ThinkSystem SE350 Edge AI platform and SAS Event Stream Processing software, to let customers analyze streaming data, uncover hidden insights with AI and make real-time, intelligent decisions.
“This is a robust platform for deploying and managing edge workloads,” Daigle said.
Another collaboration with a well-known vendor is an Edge-to-Core data management solution with NetApp.
“Edge AI increases the level of security when you can keep data locally,” Daigle noted. “This solution starts with an affordable small pod. You don’t have to use a 6U system. This really brings down the cost of entry, so that a midsize enterprise or even a smaller business could stand up an AI training solution, and then later add more compute and storage as needed.”
A retail loss prevention solution comes from a collaboration with Irish vendor Everseen, which provides visual AI technology for shaping business processes. It identifies scan errors like ticket switching, double-or mis-scanning in real-time using computer vision technology. It also improves profitability with AI capabilities, enabling cost savings and store transformation.
Finally, a partnership with cnvrg.io provides a scalable MLOps [machine learning operations] solution, a solution for data scientists that runs on Lenovo AI-Ready Servers with Intel Xeon Scalable Processors with built-in AI acceleration.
“We really see this as a robust enterprise grade platform for deploying managed AI workloads inside your data centre,” Daigle said. “This is a horizontal platform, which can be used in any industry, and which is packaged up and ready to deploy, predefined, validated and optimized on Lenovo hardware. That is what our channel has been asking for as well. They want to add the last mile around it and provide some of the skills, but they don’t have the bandwidth to validate startups themselves. This is very consumable for our channel, and part of our channel-first strategy.”
“Creating a successful AI business case involves understanding how you can apply technologies to each customer’s individual business,” Tease added. “That’s where the intimacy that channel partners have with their customers’ business is very important. It’s even more important in AI than in other use cases.”
Lenovo is also stressing that these five packaged AI solutions are just the beginning.
“We are announcing five solutions today, and a number of others will come out this year as well,” Daigle said. “This is the kickoff of our AI partner ecosystem.”