This article is more than 1 year old

HPE's GreenLake remade with fresh set of cloud services as biz starts move to aaS future

'Ezmeral' brand added to container platform and machine learning stuff

HPE is going public with a series of enhancements for its GreenLake Cloud Services, an on-premises managed computing service.

"We take our technology, we take the hardware, we take software – it might be third-party or our software – with our HPEFS financing arm, and with our HPE Pointnext services, and we bring that together and we put that into our customer's environment," said Keith White, general manager of HPE's GreenLake Cloud Services, in an interview with The Register.

"Then we meter every hour, on the hour, for the amount of CPUs or gigabytes or VMs the customer is using. And then we bill them for what they use. Then on the backend, we manage all of that for them."

The GreenLake augmentation brings customers an expanded set of menu items from the self-service HPE GreenLake Central platform, which provides a point-and-click interface for deploying machine learning tasks, containers, VMs, storage, compute, data protection, and networking options, as well as options for learning about pricing and making service trial requests.

The refurbishment brings with it a new software brand, HPE Ezmeral, which describes a suite of software for running, managing, and securing customers' applications and data. The new name means that the HPE Container Platform will now be known as the HPE Ezmeral Container Platform and HPE ML Ops becomes HPE Ezmeral ML Ops.

White describes the latter service as an on-prem version of ML Ops. "It helps data scientists get up to speed and running really quick," he said. "And we basically allow them to easily and seamlessly move models from pilot to production."

HPE has also portioned GreenLake cloud services out into pre-packaged, pre-configured building blocks, to allow customers to place orders for these use-case-tuned architectures and get them up and running on-site in as few as two weeks.

"What we're providing is basically 17 pre-integrated configurations that are all workload optimized that are available for these services from the factory direct for HPE GreenLake," White said, likening the components to Lego bricks or T-shirt sizes.

White told us the announcements represent a significant milestone for HPE. "Last year, Antonio Neri, our CEO, announced that we were going to move all of our offerings to as-a-service in the next three years," he explained. "And so this is really sort of that first implementation, first instantiation of a new generation of cloud services for GreenLake."

GreenLake differs from public cloud providers like AWS, Azure, and Google Cloud in that it doesn't charge for data egress.

"There's no data egress charge since it's on-site," said White, adding that keeping data local improves latency, which helps machine learning and AI algorithms.

White said HPE has found that its customers want to get out of the data center business. "They don't want to have to pour their own concrete or figure out how many megawatts they're going to get to their environment or what the cooling system is," he said. "They want to offload that to a colo like an Equinix or a CyrusOne, and they want to have this cloud-like experience where they can have it be all self-serve and automated, so it's all managed for them on the backend." ®

More about

More about

More about

TIP US OFF

Send us news


Other stories you might like