Everyone’s moved to the public cloud, right? No, not quite.
Sure, many organizations have moved to the cloud for application development and a place to run applications. And why not, since the benefits can include faster time to market, greater efficiencies, increased scalability and lower costs.
Yet many organizations have too many IT systems and processes to “lift and shift” them to the cloud all at once. Instead, their journey to the cloud will likely take months or even years.
In the meantime, some are adopting on-premises clouds. This approach gives them dedicated, bare metal servers, or servers that can be set up with cloud services and capabilities.
One popular approach to an on-premises cloud is Google GDC Virtual. Formerly known as Google Anthos on-prem and bare metal, this solution extends Google’s cloud capabilities and services to an organization’s on-prem data center.
Your customers can use Google GDC Virtual to run new, modernized applications, bring in AI and machine learning workloads, and modernize on-premises applications.
All this should be especially interesting to your customers if they already use the Google Distributed Cloud (GDC). This portfolio of products now includes GDC Virtual, extending Google’s cloud infrastructure and services to the edge and corporate data centers.
More help is here now from Supermicro SuperBlade servers powered by AMD EPYC processors. They’re ideal for managing cloud-native workloads. And for connecting to the wealth of services the Google Cloud Platform provides.
These servers include a bare metal option that delivers many cloud benefits to self-managed Supermicro SuperBlade servers. This offers your customers Bare Metal as a Service (BMaaS) for workloads that include AI inferencing, visual computing, big data and high-performance computing (HPC).
Why on-prem cloud?
With the public cloud such a popular, common solution, why might your customers prefer to run an on-prem cloud? The reasons include:
- Data security, compliance and sovereignty requirements. For example, privacy regulations may prohibit your customer from running an application in the public cloud.
- Monolithic application design. Some legacy application architectures don’t align with cloud pricing models.
- Demand for networking with very low latency. Highly transactional systems, such as those used by banks, benefit from being physically close to their users, data and next-hop processors in the application flow.
- Protect legacy investments: Your customer may have already spent a small fortune on on-prem servers, networking gear and storage devices. For them, shifting from CapEx to OpEx—normally one of the big benefits of moving to the cloud—may not be an option.
Using GDC Virtual, your customers can deploy both traditional and cloud-native apps. A single GDC Virtual cluster can support deployments across multiple cloud platforms, including not only Google Cloud, but also AWS and Microsoft Azure.
Super benes
If all this sounds like a good option for your customers, you should also consider Supermicro servers. They’re ideal for managing cloud-native workloads when used as control panel nodes and worker nodes to create a GDC Virtual hybrid cluster.
Here are some of the main benefits your customers can enjoy by using Supermicro SuperBlade servers powered by AMD EPYC processors:
- Hardware-agnostic: Your customers can leverage existing on-prem SuperBlade servers to drive data-center efficiency.
- No hypervisor layer overhead: Deploying GDC Virtual on SuperBlade reduces complexity.
- Rapid deployment: GDC Virtual enables rapid cloud-native application development and delivery. So both developers and dev-ops teams can benefit from increased productivity.
- Easy manageability: SuperBlade manageability, coupled with GDC Virtual management, enables increased operational efficiency. A dashboard lets you monitor what’s going on.
Under the hood
Supermicro SuperBlade servers are powered by AMD EPYC 7003 Series processors with AMD 3D V-Cache tech. These CPUs, built around AMD’s “Zen 3” core, contain up to 64 cores per socket.
Supermicro offers three AMD-powered SuperBlade models: SAS, SATA and GPU-accelerated. These can be mixed in a single 8U enclosure, a feature SMC calls “private cloud in a box.” Each server supports up to 40 single-width GPUs or 20 double-width GPUs.
Each server also contains at least one Chassis Management Module (CMM). This lets sys admins remotely manage and monitor server blades, power supplies, cooling fans and networking switches.
Another Supermicro SuperBlade feature is SuperCloud Composer (SCC). It provides a unified dashboard for administering software-defined data centers.
Have customers who want the benefits of the cloud, but without moving to the cloud? Suggest that they adopt an on-premises cloud. And tell them how they can do that by running Google GDC on Supermicro SuperBlade servers powered by AMD EPYC processors.
Do more:
- Read a solution brief: Google Anthos excels on Supermicro SuperBlade (PDF).
- Check out Supermicro SuperBlade servers.
- Explore AMD EPYC server processors.