HPE Goes Vertical with AI
HPE today unveiled new offerings aimed at getting customers up and running with artificial intelligence solutions, including an anomaly detection system for industrial equipment and a new GPU-equipped server cluster designed for deep learning workloads.
Despite the potential for AI to add nearly $16 trillion to the world’s GDP by improving labor productivity and consumer demand (PwC), only about 4% of companies have actually figured out how to take advantage of AI’s emerging capabilities (Gartner), HPE says in a press release today.
“Global tech giants are investing heavily in AI, but the majority of enterprises are struggling both with finding viable AI use cases and with building technology environments that support their AI workloads,” says Beena Ammanath, global vice president of AI at HPE Pointnext, the IT services group with 25,000 consultants that HPE launched a year ago. “As a result, the gap between leaders and laggards is widening.”
HPE aims to close the gap through a combination of tailored vertical offering, consulting and training services, and HPC hardware designed to run emerging deep learning workloads.
On the vertical front, HPE announced Digital Prescriptive Maintenance, which combines software frameworks, development work, and implementation services in one handy SKU aimed at industrial equipment operators who will spend princely sums to avoid paying the kings’ ransom that comes due when unscheduled downtime rears its ugly head.
According to the company formerly known as Hewlett-Packard, the offering brings an assortment of technologies, reference architectures from HPE and partners, and consulting services to bear on building systems that can not only predict when a given piece of equipment is not running as it should, but actually prescribe what actions its own should to take.
Per HPE: “The solution captures all relevant data sources in the enterprise, including real-time and batch data from IoT devices, data centers and the cloud. Based on both supervised learning for failure prediction and unsupervised learning for anomaly detection, HPE Digital Prescriptive Maintenance prescribes and automates actions to prevent industrial equipment failure and optimize its productivity.”
The Digital Prescriptive Maintenance offering – the first in a series of vertical AI offerings expected from HPE — is available in Europe now, and will be available around the world by June.
The IT giant also unveiled the HPE Apollo 6500 Gen10 System for training deep learning models. HPE says the new system, which will become available in May, can support up to eight NVIDIA Tesla V100 GPUs, each connected via the NVIDIA NVLink interconnect, which the company says is 10x faster than the PCIe Gen3 interconnect. For more details on the new system, see HPCwire’s story today.
The company also announced that it has struck a deal to resell WekaIO‘s file storage software, dubbed WekaIO MATRIX, together with its existing Lustre-based storage solutions. HPE says WekaIO’s flash-optimized file system can support the demanding I/O required for large-scale data analytics and AI environments.
“The HPE Apollo 6500 Gen10 System is purpose-built to enable organizations of all sizes realize the benefits of deep learning faster than ever before,” Pankaj Goyal, HPE’s vice president of hybrid IT strategy and AI, says in a press release. “And with WekaIO’s flash-optimized parallel file system, HPE now provides the required throughput for compute-intensive low-latency workloads.”
HPE also announced its Artificial Intelligence Transformation Workshop, a new consulting offering whereby HPE Pointnext experts help customers to get started with AI. Figuring out where to apply AI and big data technology can be a challenge, and HPE is hoping to help customers by evolving their data strategies and identifying AI use cases.
Lastly, HPE announced its Deep Learning Performance Guide, a follow-on to last year’s release of the HPE Deep Learning Cookbook. The company says that the guide makes recommendations for the right hardware and software stack for given analytic or AI workloads, including identifying bottlenecks in hardware.
“Customers pursuing deep learning projects face a variety of challenges including a lack of mature use case and technology capabilities that can compromise time to value, performance and efficiency,” said Steve Conway, a senior vice president with Hyperion Research and a contributor to Datanami. “HPE’s domain expertise, services, technologies and engineering ties to ecosystem partners promise to play an important role in driving AI adoption into enterprises in the next few years.”