HPE continues to develop its hardware-as-a-service (HaaS) offering, brought together under the HPE GreenLake. It did well if we rely on its annual results published at the end of November. On a turnover of $28.5 billion (+3% over one year), the still recent HaaS activity represents $1 billion, up 17%.
As a reminder, GreenLake is the hybrid multicloud offer, public or private, managed by HPE and for which users pay per use. During its European Discover 2022 conference, held in early December in Frankfurt (Germany), the supplier claimed more than 900 partners on GreenLake and 65,000 customers, with a retention rate of 96%.
In GreenLake, HPE delivers advice according to usage, provides deployment and administration, with SLAs as the key. “One of the most interesting use cases is that of developers who program on the public cloud in a DevOps environment and then bring their applications back to their company’s private cloud for compliance issues,” said Antonio Neri, President and CEO of HPE.
“Today, half of companies use the hybrid cloud. We estimate that this proportion will increase to 70% within a few years,” he adds.
New features thanks to partnerships
HPE considers the number of partners to be critical as they drive revenue from its marketplace. In GreenLake, HPE does not intend to systematically take on the role of publisher, hence the importance of having reliable integrator partners to ensure SLAs (Service-Level Agreement) of the applications that it sells on its market place.
The latest evolution of GreenLake thus supports EKS (Elastic Kubernetes Service), the AWS signed version of Kubernetes. “Many of our customers want to use EKS in their private cloud, not public, without having to change anything. Kubernetes has become a standard, but there are different specificities and implementations depending on its variations,” says Bryan Thompson, VP of GreenLake Cloud Product Management.
DataFabric is another notable step forward: “You have to think of DataFabric as storage “as-a-service”. I can have S3 type object storage, NFS for files, or Azure storage. DataFabric provides a single view and allows data to be collected and processed wherever it comes from”, insists Bryan Thompson.
Finally, GreenLake offers more specialized instances depending on the uses. For example, the focus will be on memory for configurations intended for analytics in RAM, such as SAP S4/HANA, on computing for applications with NoSQL databases, or even on molecular dynamics. And HPE intends to continue its verticalization strategy, already well underway in the medical sector, with new partnerships.
All this does not prevent HPE from developing its own solutions, particularly in terms of artificial intelligence. And more particularly concerning the machine learning (ML), whose supplier foresees the evolution towards Swarm learning.
Inspired by biology, and based on a blockchain, Swarm learning consists in a way of decentralized ML: the data is not moved, it remains at the edge of the network (edge). They are therefore not shared, only the AI models are: “this is an operation that we have already implemented with the exchange of information between health professionals on lung x-rays”, illustrates Antonio Neri .
We want to give thanks to the writer of this write-up for this remarkable material
HPE multiplies the features behind GreenLake
Visit our social media accounts as well as other pages related to themhttps://www.ai-magazine.com/related-pages/