Search
Homepage This page's url is: -crn- Rankings and Research Companies Channelcast Marketing Matters CRNtv Events WOTC Jobs HPE Discover 2019 News Cisco Wi-Fi 6 Newsroom Dell Technologies Newsroom Hitachi Vantara Newsroom HP Reinvent Newsroom Lenovo Newsroom Nutanix Newsroom Cisco Live Newsroom HPE Zone Tech Provider Zone

HPE Accelerates Machine Learning With ML Ops Container Software

‘Partners need to build out their (machine learning and AI) practices, bench strength and expertise in this area so they can help customers in the next phase of the journey,’ says HPE BlueData Vice President of Marketing Jason Schroedl.

Hewlett Packard Enterprise is aiming to accelerate the rollout of full production environment machine learning solutions with its new ML Ops Docker container-based software platform.

In fact, HPE said it expects to reduce the time it takes to implement full lifecycle machine learning software deployments from months to just days with ML Ops.

[Related: 5 Things You Need To Know About HPE’s Machine-Learning BlueData Platform]

HPE ML Ops is an end-to-end-based full lifecycle platform that provides software to build, train, deploy and monitor machine learning solutions. It is a multi-user, collaborative environment with Git version control and file system integration with data repositories.

The new ML Ops platform is the next wave in the AI and machine learning market moving partners from pilots and proof of concepts to full “production” deployments for multiple projects, said Jason Schroedl, vice president of marketing for HPE’s BlueData AI-based big data software unit.

“This is operationalizing machine learning, moving it into (a full) production environment that can be deployed for multiple projects throughout the organization,” said Schroedl.”It takes us from a handful of data scientists and analysts building machine learning models with an artisan approach to a more industrialized approach that can be deployed at scale.”

 HPE ML Ops effectively brings DevOps agility to the exploding machine learning marketplace, said Schroedl. That translates into a “gold rush opportunity” for solution providers to build out machine learning practices, he said.

“Every enterprise is doing something in this area,” said Schroedl. “Partners need to build out their practices, bench strength and expertise in this area so they can help customers in the next phase of the journey. The opportunity for channel partners is to leverage technology and solutions like we are bringing to bear and then build out their practice as customers look to operationalize their machine learning initiatives.”

The new HPE ML Ops solution extends the capabilities of the BlueData EPIC big data software platform, which enables data science teams to quickly spin up containerized environments for distributed AI / ML and analytics solutions in hybrid cloud environments.

A key difference for the new ML Ops solution with BlueData, in fact, is its ability to run in a hybrid cloud environment – either on premise or in a public cloud environment, said Schroedl. “This is a hybrid cloud solution that can be run on prem or in the cloud or some combination of the two,” he said. “You can run this on AWS, Azure, GCP. That is one of the unique capabilities we bring to bear.”

HPE acquired BlueData as part of a no holds barred AI and machine learning market share grab that also includes the $1.3 billion acquisition of Cray which is expected to be finalized this quarter.

The HPE BlueData team is in the midst of a partner recruitment offensive aimed at driving high margin, recurring revenue big data opportunities for the channel.

“We are building out our go to market team, our sales team and channel partner relationships in target geos (geographies),” said Schroedl.

Nanda Vijaydev, distinguished technologist and lead data scientist for BlueData at HPE, said there is no other on premise, hybrid cloud container based offering that can match HPE ML Ops. “This goes through the full lifecycle of machine learning with development, training, deployment and monitoring- all using containers,” she said. “It is the equivalent of DevOps. That’s why we call it ML Ops (Machine Learning Optimization).”

The mainstay of the platform is “quick provisioning, collaboration, and the ability to deploy machine learning models in a containerized, scalable deployment with end point security and monitoring,” said Vijaydev. “It is a holistic, end to end complete lifecycle offering in a containerized environment which is what BlueData is famous for.”

The new HPE ML Ops lifecycle software provides a wide choice of tools and framworks including integration with Docker Endpoint Wrapper, Python ML and DL Toolkit, Spark Training, TensorFlow ModelServer, and Jupyter Notebook open source document sharing.

HPE ML Ops’ diverse ecosystem of tools along with its complete end to end lifecycle approach to machine learning is unique, said Vijaydev. “Most of the cloud solutions have a very fragmented approach to the lifecycle,” she said. “BlueData is a single pane of glass- a single control panel for same cloud, multi-cloud and hybrid.”That diverse BlueData ecosystem prevents cloud and vendor lock in, she said.

Nalit Patel, CEO of All Solutions, a Livingston, N.J-based solution provider, said the AI and machine learning sales charge is yet another testament to HPE’s innovation prowess.

Patel credited HPE CEO Antonio Neri, who has competed 15 acquisitions, with helping partners develop innovative solutions for customers in lucrative new markets like AI and machine learning.

“HPE is providing us with an end to end ecosystem for providing innovative solutions to customers,” he said. “It’s a holistic platform with continuity so we don’t need to go to 20 different vendors. HPE is investing billions in research and development and acquisitions.”

Back to Top

Video

 

sponsored resources