One week after IBM Corp. previewed its currently running Think 2018 conference with a series of artificial intelligence-related announcements, Hewlett-Packard Enterprise Co. is following with its own take on the hottest trend in information technology.

The company today announced hardware, vertical-market software and training services aimed at getting customers up to speed on machine and deep learning.

The hardware piece is the HPE Apollo 6500 Gen10, which is described as a high-performance computing system purpose-built for deep learning that conducts model training three times faster than previous generations.

Model training is a key part of the machine learning process that creates a plan of action for algorithms to use in analyzing large amounts of data. Deep learning is a type of machine learning that attempts to mimic the functions of the human brain. The system is similar to one that IBM recently rolled out based upon its new Power9 processors.

Outfitted with Intel Xeon Scalable processors with up to 28 cores operating at 3-gigahertz frequency, along with eight graphics processing units per server, the system delivers up to 125 trillion floating-point operations per second for single-precision compute tasks. It comes with either a PCI Express backplane or Nvidia Corp.’s NVLink 2 interconnects for high-speed GPU-to-GPU communications.

High-bandwidth, low-latency networking adapters are tightly coupled with the GPU accelerators for maximum bandwidth utilization, the company said. Each system supports up to 3 terabytes of memory and 16 Serial Attached SCSI or Serial AT Attachment solid state drives, of which four can be NVM Express drives. “The HPE Apollo 6500 Gen10 System best addresses the most important step of training the deep learning model,” HPE said in an announcement.

HPE Digital Prescriptive Maintenance is described as the first in a series of AI-optimized industry solutions. It automates problem prevention through predictive maintenance, thereby increasing the lifespan of industrial equipment while reducing downtime. “While predictive maintenance detects when industrial equipment is likely to fail, prescriptive maintenance predicts, suggests and automates the right action to fix the problem before it causes harm,” the company said.

The package combines educational and implementation services from HPE’s PointNext services group, along with machine learning and data analytics software from HPE and its partners on top of an internet of things-driven data lake based on HPE hadware. The service incorporates real-time and batch data, machine learning, analytics-based anomaly detection, data protection in the event of equipment failure, guided maintenance steps and automated workflows, either at the edge or at the center of the network.

Finally, the HPE Artificial Intelligence Transformation Workshop is a new, one-day customized introduction to AI for business leaders based upon use cases aligned to the company’s business. Participants can expect to “discover the areas that need attention and create a high-level plan with opportunities, obstacles, and critical success factors that are specific to your needs,” the company said. The plan includes recommendations for how to move from experimentation to production.

Pricing wasn’t announced.

Image: Clever Cogs! via photopin (license)