HPE hosted their Discover event in Las Vegas from June 20 through 22. The overall theme of the conference was the hyper-connected enterprise and the impact of the digital transformation.
Antonio Neri, CEO, chaired the opening session focusing on how digital transformation’s impact is accelerating business time to value through technology and people, all underscored by the new economics of the consumption economy. HPE made several announcements, the biggest of which were focused on what they referred to as the “Intelligent Edge.”
HPE proposes that enterprises cannot move all data that originates at the edge to the cloud or the centralized data center; there is too much data and there is not enough bandwidth to service response times. Thus data will be processed first at the edge to service the connected economy. To this end they announced a $4 Billion investment over the next 4 years to accelerate development of the Intelligent Edge. While this likely will be Aruba Networks investments (where they had major announcements, and is not covered in our commentary) they also announced the new Edgeline Converged Edge Systems (CES)–hardened systems that incorporate Xeon cores (64), integrated accelerators (aka GPU), storage (up to 48TB solid state), with the ability to integrate operational technology (OT) and run everything from full function SAP, SparkRecognition, Linux, VMware and a long list of other platforms. This contrasts to other vendor strategies we have seen for “skinnied-down” edge systems with specialized software. The question is why? HPE’s vision is to address the operational and information technology complexity that is prevalent in manufacturing, mining, auto systems, energy, etc., and recognizes that we can now combine OT and IT into the same box for real time processing and efficiencies. Expect HPE to have a regular cadence of new offerings and invest in the required bus dev and sales efforts to penetrate this new space. To underscore the importance of these investments, when Neri was asked about growth and what the company will look like 5 years from now, he expected “the most growth from the edge.”
For our readers who focus on the data center, HPE outlined their HPE / AI in support of the edge to core strategy (see Figure 1) that includes a data lake for hot, warm and cold data and support for fast data analytics (Hadoop / Spark) running on Apollo 4000 data storage device. The data storage is configured using several ISVs for the storage personalities, including Weka.IO and Qumulo for scale-out file, and Scality for object storage. Depending on the workload; streaming, training or analysis: the workloads are serviced with the various compute versions of Apollo with CPU / GPU or Synergy. None of these are new products or offerings, but in light of the hyper-connected message, an important component of their offerings for the connected enterprise.
No HPE conference would be complete without an appearance of “The Machine.” While there were no Machine announcements, HPE did announce a Memory Driven Sandbox for developers which runs on Superdome. Antonio Neri pointed out that we will continue to see technologies developed in conjunction with The Machine- show up in other places, such as photonics capability or persistent memory in Synergy. Neri’s comment also unscored his commitment to presenting HPE as an “engineering-based company.” He reminded the press and analysts that HPE developed its own silicon processing power delivered in Aruba, The Machine and Superdome. It was also evident in the Hewlett Packard Lab displays on the show floor.
Lastly we take note of the Greenlake Hybrid Cloud as a service consumption model announcement. On premises is AzureStack, connected to Azure or AWS off premise and managed through OneSphere (see Figure 2). The offering is metered with not to exceed cost controls plus continuous compliance monitoring. HPE delivers this though their PointNext team of Cloud Technology Partners and Red Pixie, two companies with cloud deployment specializations that they acquired in the last 12 months.
HPE is all-in on the Connected Edge. They are well established in data center and HPC but will need to drive further into AI technologies (Deep Learning) to capture the new market opportunity presented by Edge computing. We believe there are still organizational collaboration challenges within HPE, but we note that those challenges have significantly lessened in the past year, some of this evidenced with their Edge to Core offerings. Scality is reviewed in Evaluator Group Object Coverage. Weka.IO and Qumulo will be reviewed in the fall in the Scale-out File coverage.