IBM has been on a roll lately with POWER Systems. Principal analyst and colleague Patrick Moorhead wrote aboutGoogle deploying POWER9 announced at the OpenPOWER Summit and last week I wrote about IBM expanding its platform HCI footprint with Nutanix. As another output of a long partnership, IBM and Red Hat recently announced the availability of IBM Cloud Private on Red Hat OpenShift/ Red Hat Enterprise Linux (RHEL). Additionally, IBM recently announced the availability of PowerAI for Red Hat Enterprise Linux (RHEL) running on IBM POWER9 platforms. What does this mean and how does this impact the enterprise IT organization?
What is IBM Cloud Private and OpenShift?
Before getting into the details of the partnership, a little refresher on IBM Cloud Private and Red Hat OpenShift.
Cloud Private is IBM’s private cloud platform that enables enterprise IT to employ a hybrid cloud environment on both x86 and POWER platforms. IBM sees Cloud Private as addressing three enterprise IT needs:
Source: Moor Insights & Strategy
IBM’s value prop is essentially save money on legacy app support, securely integrate with third parties for implementations such as Blockchain and simply develop twelve-factor cloud-native applications in a microservices architecture. Important to note in Cloud Private is its ability to run in both POWER and x86 environments.
What this partnership enables
As previously mentioned, the partnership extends IBM Cloud Private to Red Hat OpenShift. So, enterprise IT organizations familiar with the Red Hat tools can more simply deploy a cloud environment that brings all its data and apps together in a single console. Legacy line of business (LoB) applications can be deployed and managed alongside native cloud applications. IBM middleware can be deployed in any OpenShift environment.
This partnership also allows a simpler, more secure interface to the power of IBM Cloud Services. The seamless integration from IBM Cloud Private should allow IT organizations to quickly enable services that would normally take months to deploy such as Artificial Intelligence (AI), Blockchain and other platforms.
Source: Moor Insights & Strategy
PowerAI and RHEL brings Deep Learning to the enterprise
Somewhat hidden in the news of the IBM – Red Hat announcement is what may be the most interesting bit of news. That is, the availability of PowerAI for RHEL 7.5 on the recently updated Power System AC922 server platform.
PowerAI is IBM’s packaging and delivery of performance-tuned frameworks for deep learning such as Tensorflow, Kerras, and Caffe. This should lead to simplified deployment of frameworks, quicker development time and shorter training times. This is the beginning of democratizing deep learning for the enterprise. You can find more on PowerAI here by Patrick Moorhead.
The IBM POWER System AC922 is the building block of PowerAI. As previously mentioned, this is based on the IBM POWER9 architecture. Why does this matter? In an acronym, I/O. POWER9 has native support for both PCIe Gen4, NVLink 2.0 and CAPI 2.0. Both of these allow for greater I/O capacity and bandwidth. Moreover, what that means to a workload like deep learning is the ability to move more data between (more) storage and compute much faster. This leads to a big decrease in learning time. To an enterprise IT organization, that means faster customer insights, greater efficiencies in manufacturing and a lot of other benefits that drive differentiation from competitors.
What this means for Enterprise IT
There are a few ways this partnership benefits the enterprise IT organization. One of the more obvious benefits is the tighter integration of applications and data, both legacy and cloud-native. Enterprise IT organizations that have gone through the pains of trying to integrate legacy data with newer applications can more easily take advantage of IBMs (and open source) middleware to achieve greater efficiencies.
This partnership also allows enterprise IT to more quickly enable a greater catalog of services to business units looking to gain competitive advantages in the marketplace through IBM Cloud Services.
Perhaps the biggest benefit to enterprise IT is the availability of OpenAI on RHEL. I believe this begins the democratization of AI for the enterprise. This partnership attempts to remove the biggest barriers to adoption by simplifying the deployment and tuning of Deep Learning frameworks.
How this benefits IBM and Red Hat
IBM can extend the reach of its cloud services to enterprise IT organizations running Red Hat OpenShift. I believe those organizations will quickly be able to understand the real benefits associated with Cloud Private and Cloud Services.
The benefit to Red Hat is maybe a little less obvious, but equally significant. Red Hat’s support for IBM Cloud Private and (by extension) Cloud Services opens the addressable market for OpenShift and enables a new set of differentiated capabilities. In an ever-increasing competitive hybrid cloud management space, this sets Red Hat apart.
On the AI front, I believe this partnership further sets IBM apart as the leader and introduces Red Hat into the discussion for good measure. This could be a partnership that many try to catch for some time.
The partnership between IBM and Red Hat has always been strong, and in many ways these solutions offerings only make sense. Red Hat has a strong offering in the developing, deploying and managing cloud-native applications with OpenShift. IBM has a best-of-breed solution in Cloud Private and PowerAI. Marrying the two can empower the enterprise IT organization and extend the datacenter footprint of both Red Hat and IBM.
However, many great technical partnerships never reach their potential because the partnerships end at technical enablement. Red Hat and IBM would be wise to develop a comprehensive go-to-market campaign that focuses on education and awareness. Cross-selling and account seeding is the first step in enabling this partnership, followed by a series of focused campaigns in targeted vertical industries and market segments.
Finally, the joint consulting services between the IBM Garage and Red Hat Consulting organizations will need to work closely in ensuring early customer success with PowerAI. Real enterprises realizing real benefits is the difference between a science project and a solution. Moreover, these organizations are going to be critical to helping enterprise IT stand up these deep learning frameworks.
I will be following this partnership closely and look forward to watching how Red Hat and IBM jointly attack the market. Look for a follow up on this as the partnership evolves.