Since the dawn of ‘the cloud’, hyperscale cloud providers have been the go-to choice for businesses in need of near-instant scalability. But over time, excessive overheads and performance trade-offs have started to take their toll.
Because of this, hyperscale cloud alternatives like bare metal server hosting have surged in popularity. But while these solutions solve cost and performance pains head on, the vast majority still don’t offer the near-instant scalability of hyperscale cloud environments.
We’ve had countless businesses reach out to us with this dilemma. Multi-vendor approaches go some way to solve the problem, but we knew a more effective solution was possible.
That’s why we developed Scalable Bare Metal (SBM) - a first of its kind product offering combining bare metal power and performance with the flexibility to scale on-demand.
“We expect SBM to become an important part of the servers.com portfolio and a go-to solution for a wide range of customers - from DevOps and infrastructure teams to companies building cloud platforms, SaaS services, high-traffic applications, data processing workloads and gaming projects. Our goal is to deliver a product that provides fast deployment, stable performance and predictable infrastructure on a global scale.”
SBM is a highly scalable bare metal cloud solution developed by servers.com that addresses the limitations of traditional dedicated server models in terms of flexibility and scalability. With SBM, bare metal servers are spun up in minutes and billed by the hour, offering on-demand compute without compromising performance or resource isolation.
SBM is designed for use alongside our Enterprise Bare Metal (EBM) servers as a source of burst capacity.
For most businesses, demand oscillates between stable periods and occasional peaks and troughs. During periods of stability, custom solutions like EBM are ideal. EBM can be optimized to your exact performance requirements to become extremely cost and resource efficient.
However, EBM can’t be scaled on-demand. To cover unexpected or temporary demand spikes you’ll still need a separate source of highly scalable compute. That’s where SBM comes into the picture. While, at the hardware level, SBM is pure bare metal hosting (just like EBM), it’s significantly more scalable with 15-minute provisioning and hourly contracts.
In this way, SBM functions as a ‘bolt on’ to EBM, providing a source of highly scalable bare metal compute that you can tap into as and when you need it.
What makes SBM so scalable is its standardized setup. All SBM servers available for order are pre-provisioned and pre-configured in fixed hardware configurations (known as flavors). This significantly expedites spin up times and allows users to easily grow their infrastructure by adding additional identical nodes with predictable specifications.
There are four base configurations available, designed for varying types of workloads from entry level compute to performance-intensive tasks and large-scale application hosting.
| Base configuration | CPU | RAM | Storage |
|---|---|---|---|
| Entry line | 4564P | 192 GB | 2 x 960 GB NVMe |
| Performance line | 9474F | 384 GB | 2 x 3.84 TB NVMe |
| Business line | 9354P | 256 GB | 2 x 1.92 TB NVMe |
| 9554P | 512 GB | 2 x 3.84 TB NVMe |
Flavors are based on market trends, customer demand, and vendor benchmarks to create optimal CPU and component combinations.
SBM and EBM are designed to support CPU, not GPU, workloads. If you also have GPU workloads that need hosting, that’s where our AI Compute (AIC) service comes into play.
AIC offers highly optimized GPU dedicated servers on long-term agreements. And, crucially, it works seamlessly alongside SBM and EBM as part of our hybrid bare metal cloud offering.
By combining these solutions, you can host both GPU and CPU workloads on one private network and avoid the operational complexity of juggling multiple vendors.
Historically, scalable infrastructure has been closely associated with hyperscale cloud services. These services rely on hardware abstraction to create a ‘virtualized’ infrastructure.
Each server is equipped with a hypervisor which splits the server’s resources between multiple VMs. Each VM is then assigned to a different user and the operating system and application placed on top.
Virtualized environments are highly scalable because VMs can be duplicated near instantly and assigned resources dynamically. However, the sharing of resources can result in performance tradeoffs from resource contention. And in hyperscale cloud environments, the added flexibility comes with a significant cost overhead.
By contrast, SBM provides on-demand scalability and the full tenant isolation of bare metal, so there’s no hypervisor overhead or risk of resource contention impacting server performance. This makes SBM ideal if you need to scale quickly but also have strict requirements around security, latency and workload predictability.
| SBM | VMs | |
|---|---|---|
| Definition | Physical servers dedicated to a single tenant | Physical servers shared between multiple tenants |
| Scalability | Provision on demand in minutes | Provision on demand in minutes |
| Performance | High performance with no virtualization overhead | Potential for performance tradeoffs from resource contention |
| Resources | Dedicated resource allocation | Shared resources |
| Cost | Opportunity for cost optimization when combined with EBM | On-demand premium |
| Complexity | Simple deployment and transparent billing | Complex billing that can lead to vendor lock-in |
If you’re looking for a bare metal infrastructure solution that you can burst into to accommodate temporary spikes in demand, SBM offers several key benefits, including:
Provisioned in minutes and billed by the hour, SBM is designed for on-demand infrastructure scaling during peak loads. Combining scalability with bare metal power and performance, SBM is ideal for workloads that are likely to present some volatility. Examples include high-traffic web applications, real-time data processing workflows and game server hosting.
Based on pure bare metal hardware (not VMs), SBM offers full hardware isolation, consistent network infrastructure and high-performance compute without compromising on scalability. This makes it an ideal solution for security sensitive environments and CI/CD pipelines that need a source of isolated, predictable compute resource available on-demand.
With fixed flavors optimized for specific use cases, SBM can be deployed in 15 minutes directly from the servers.com customer portal. Because SBM is billed by the hour, there’s no long-term contracts or financial commitments. And as a pure, bare metal solution there’s also no proprietary element or risk of vendor lock-in associated with using the product.
SBM is incredibly easy to utilize as part of our hybrid bare metal cloud offering. SBM has been designed to be completely interoperable with both EBM and our AI Compute (AIC) services. This means you can combine a highly optimized baseline infrastructure in EBM, tap into SBM when you need an extra burst of scalability and host AI workloads – all on the same network.
As businesses grow tired of the excessive overheads and performance trade-offs that come with virtualized infrastructure environments, more are seeking hyperscale cloud alternatives that still deliver on flexibility and scalability.
By reimagining how bare metal hosting is delivered, SBM provides a compelling alternative where full resource isolation and high performance meet fast deployments and hourly billing. Pure bare metal hardware with on-demand scalability.
Whether you're managing volatile workloads or building a hybrid setup, SBM is a simple way to access flexible, scalable compute on your terms.
SBM provides the same raw performance and isolation as traditional bare metal but with near-instant scalability. SBM servers spin up in minutes and are billed by the hour.
Hyperscale cloud providers offer instant scalability but at the cost of high overheads, complex billing, and resource contention due to virtualization. SBM delivers comparable scalability while keeping the benefits of bare metal: predictable performance, full hardware isolation, and transparent hourly billing.
SBM is designed to complement Enterprise Bare Metal (EBM) by providing burst capacity. Many customers use EBM for stable, predictable workloads and add SBM to handle temporary spikes in demand. However, depending on your use case, SBM can also serve as the primary infrastructure for workloads with highly variable demands.
SBM is ideal for workloads that experience unpredictable or volatile demand, such as:
No. SBM is optimized for CPU-based workloads. For GPU compute, you can use our AI Compute (AIC) service, which integrates seamlessly with both SBM and EBM in a hybrid bare metal cloud environment.
All SBM servers are pre-provisioned and can be deployed in around 15 minutes directly from the servers.com portal.
Flavors are standardized hardware configurations designed around customer demand and market benchmarks. By fixing the configurations, provisioning is faster, billing is simpler, and scaling is more predictable.
SBM is billed hourly with no long-term contracts. This makes it a cost-effective option for temporary demand spikes and test environments.
Yes. Unlike hyperscale cloud providers, SBM runs on pure bare metal hardware with no proprietary elements, so you avoid vendor lock-in and maintain infrastructure flexibility.
Yes. SBM is designed to work seamlessly with Enterprise Bare Metal (EBM) and AI Compute (AIC), allowing you to build hybrid environments with baseline infrastructure, burst scalability, and GPU compute – all connected on the same private network.
Frances is proficient in taking complex information and turning it into engaging, digestible content that readers can enjoy. Whether it's a detailed report or a point-of-view piece, she loves using language to inform, entertain and provide value to readers.
