Taking Innovation to New and Unexpected Levels at VMworld 2018
Today’s opening VMworld keynote showcased a variety of new products as well as updates to our core product portfolio that build on our multi-year any cloud, any application, any device strategy. You can read all about those updates in my VMworld announcement summary and analysis post. In addition, the opening keynote also included a number of highly innovative and likely unexpected surprises for many in the audience. We continue to accelerate the pace in which we bring new VMware innovations to market, and today we offered a number of compelling proof points:
- Tech Preview: AWS RDS Running Locally on VMware SDDC
- Tech Preview: Project Dimension
- Tech Preview: Project Magna—The Self-Driving Data Center
- Tech Preview: Virtualization on 64-Bit ARM
- Project Concord
AWS RDS Running Locally on VMware SDDC
Last year at VMworld U.S., we ended the day two keynote by questioning the possibilities of the types of apps and services that could potentially run on our platform. Over the past year, it has become even more apparent to both our customers and cloud provider partners that there are many times where it just makes more sense to move a cloud service to the data center instead of moving a large data set to the cloud. In essence, cloud does not have to be a statement of where, but more of an operational model.
If it makes more business sense to run a subset of cloud services in a data center or at an edge site, we want to help you and our cloud partners to achieve that. In many ways, you can think of cloud services as a new generation of application platforms. To that end, virtualization becomes a natural way to run an aggregate of cloud services, containers, traditional applications, and whatever is necessary to meet business requirements.
Machine learning, inference, IoT and even augmented reality drive the need to bring application and data services closer to where data is either created or consumed. That is why today at VMworld, we demonstrated another phase of our strategy. We announced the tech preview of AWS RDS on VMware.
The RDS demo was very compelling in terms of developer experience and the operational consistency. The demo showed how you can define a custom availability zone on AWS and map it to an on-premises VMware vSphere cluster. At that point, the vSphere cluster is managed as just another availability zone in the AWS console.
Of course, you wouldn’t want to deploy RDS out to the data center or edge if a loss of connectivity to the AWS cloud would result in an outage. We thought that issue through, and AWS RDS on VMware will continue to operate even if there is a network disruption to the AWS cloud.
VMware admins can manage RDS on-premises using a new VMware vCenter plug-in and can monitor RDS using VMware vRealize Operations. For anything that lands in the data center or at the branch or edge, VMware admins will be able to use the tools they know and love to ensure operational consistency and remove any need for operational silos. In addition, software developers can manage RDS using the AWS console or AWS APIs or any AWS-compatible tool of choice.
Because the service is deployed locally, you can now get all the benefits of RDS and have the choice as to whether data leaves your data center. That’s bringing the velocity and scale of the cloud anywhere you need it without having to choose between one operating model or another. That is exactly how hybrid cloud should be.
Developers manage RDS from AWS console or APIs and will be able to create a customer availability zone that includes a vSphere cluster on-premises. For bringing cloud services to the Edge and on-premises, this is just the start. #vmworld
— Chris Wolf (@cswolf) August 27, 2018
Tech Preview: Project Dimension
Our customers frequently tell us that they really like the cloud model to consume VMware infrastructure, whether it’s through IBM Cloud, VMware Cloud on AWS or any one of our 4,000 cloud provider partners. At the same time, they tell us that they’d like to have a similar model for consuming infrastructure that physically resides in their data centers, branch or edge sites.
Project Dimension addresses that specific need, taking the simplicity, agility and velocity you want in cloud infrastructure-as-a-service and physically bringing it to wherever infrastructure is required. You’ll get centralized management and brokering from a SaaS-based VMware Hybrid Cloud Control Plane and have the infrastructure managed by VMware end-to-end. That includes all core software-defined data center (SDDC) components: vSphere, vSAN and NSX.
Ultimately, you get VMware Cloud Foundation as a hyperconverged appliance fully managed by VMware. You don’t have to worry about hardware and software infrastructure maintenance, just like how that’s not a concern in the public cloud. Our initial offering will be in partnership with Dell and Lenovo. For more information on Project Dimension, read Kit Colbert’s informative blog post.
Tech Preview: Project Magna—The Self-Driving Data Center
We have a deep bench of machine learning experts at VMware. As a result, we’ve churned out some very sophisticated machine learning capabilities across a range of products, including Wavefront, AppDefense and Workspace ONE Intelligence. In fact, the way in which we use machine learning in all three of those products was featured in the opening keynote.
At VMware, our machine learning strategy is focused on three primary goals:
- Smarter Customers: This entails ensuring that our solutions support the machine learning and inference projects, products and use cases our customers care about. That includes support for projects, such as Tensorflow and MXNet, and innovations with hardware partners like NVIDIA. The goal is to help you do what was once thought to be impossible: live migrate workloads connected to GPUs. Finally, we are working with a range of partners who are writing code and building solutions to help our customers accelerate innovation and enable more data-driven decisions.
- Smarter VMware: We use machine learning internally to better understand customer sentiment and feedback on our products and solutions. That helps us bring better products to market and ensure we are focusing on the right customer requirements.
- Smarter Products: Machine learning, intelligence and analytics are key design elements across our range of products. As our products grow their intelligence, you’ll continually have more time to focus on other IT needs in your organization.
— Chris Wolf (@cswolf) August 27, 2018
Project Magna is part of our smarter products initiative. It focuses on using reinforcement learning to understand an application’s needs and to continually optimize the application configuration and underlying infrastructure configuration to drive greater performance and efficiencies. The keynote demo illustrated how Project Magna could learn and understand application behavior to the point that it could model, test and then reconfigure the network to a more optimal network frame size to improve performance.
You can imagine these types of enhancements encompassing all components of the SDDC, including vSAN storage and the hypervisor configuration. Our ultimate goal is to give you a safe, highly secure and continually self-optimizing data center infrastructure with the same level of performance and quality you’d expect from a self-driving car. For more information on our machine learning initiatives, I encourage you to read the following blog posts:
- AI and the Birth of the Next Generation Data Center, by Dev Nag
- AI and Deep Learning at VMworld 2018, by Gina Rosenthal
Tech Preview: Virtualization on 64-Bit ARM
— Chris Wolf (@cswolf) August 27, 2018
Over the past few years, we have been deeply involved with customers in many edge and IoT use cases. Many of those customers asked us about bringing our virtualization technologies to 64-bit ARM, because there is a vibrant ARM ecosystem around their IoT and edge environments. In today’s keynote, we demonstrated a tech preview of ESXi on 64-bit ARM running on a windmill fARM at the edge.
The demo was shown on a Marvell MACCHIATObin, and we are also currently running the hypervisor on a SolidRun ClearCloud 8K edge gateway in our lab. It should be of note that this is not a standard VMware product. Our planned route to market is in partnership with edge device vendors. We plan to work with select embedded OEMs to scope and explore opportunities for focused, ARM-enabled products at the edge. We have no plans to support ARM in the data center at this time. If you have a compelling use case and want to move forward with virtualization on ARM in a big way, we want to hear from you! Please reach out to your VMware account team or feel free to connect with me or Daniel Beveridge on social media.
Tech Preview: Project Concord
The opening keynote wrapped up with one final project announcement: Project Concord. Project Concord is an open-sourced, highly scalable, energy-efficient decentralized trust infrastructure for digital consensus and smart contract execution. Concord can be used to power distributed trust infrastructures, including blockchains. As a leader in enterprise-grade distributed management and security products, VMware is poised to help enterprises address their blockchain goals.
Our vision for blockchain is to create an enterprise-grade service by surrounding Concord’s consensus engine with additional technologies required to enable trusted information sharing across organizational boundaries. Key elements of that enterprise blockchain vision include:
- Customer choice in deploying managed data replicas in public or private clouds;
- Developer friendly interfaces (such as Ethereum’s contract language);
- Audit and compliance capabilities;
- Enterprise-ready security and fault tolerance; and
- Integration with other VMware offerings.
With VMware digital trust infrastructure, you and your partners will soon be able to work together in confidence that your digital transactions can be trusted and verified. We are very excited about this project launch. I encourage you and your organization to get involved.
Our Work Isn’t Done!
Showcasing our emerging innovations is an important step in getting feedback on where we are and what requirements you may need us to address. Please find us and connect at or after VMworld. Everything we build is in partnership with our customers and ecosystem partners. Let’s keep innovating. Together.