What to Look For in a Bare-Metal Computing Provider

Oct. 5, 2021
A specialty bare-metal cloud provider can save customers between 45% and 80% off the price of comparable offerings from the three largest public cloud companies, according to a DCF special report, courtesy of phoenixNAP and Intel.

Last week in our special report series, we looked at common use cases for bare-metal servers and why bare-metal cloud computing is gaining momentum. This week, we continue with an overview of things to look for when choosing a bare-metal computing provider.

Choosing a bare-metal computing service provider

Get the full report.

All major cloud vendors offer bare-metal options, but most have come late to the market and their enthusiasm around bare-metal computing has been muted.

A specialty bare-metal cloud provider can save customers between 45% and 80% off the price of comparable offerings from the three largest public cloud companies.

There are good reasons for that. Cloud vendors specialize in operating highly automated hardware environments with high utilization rates enabled by virtual machines. They endeavor to make the underlying infrastructure as invisible to their customers as possible, which also gives them the flexibility to shift workloads between data centers and regions as they need to balance capacity.

Hosting dedicated hardware in data centers is neither an area of expertise for public cloud providers nor a compatible business model. For these and other reasons, licensing bare-metal computing from a major hyperscale cloud provider can cost significantly more than working with a vendor that specializes in bare-metal computing. One analysis conducted by phoenixNAP found a specialty bare-metal cloud provider can save customers between 45% and 80% off the price of comparable offerings from the three largest public cloud companies.

What to look for in a bare-metal computing provider

Given the growth statistics cited earlier, it isn’t surprising that many service providers now offer bare-metal options. The expertise needed to operate bare-metal and multitenant cloud environments differs in some fundamental ways, however. Look for a company that can demonstrate that bare-metal computing is a core competency, not an afterthought. Inquire about the company’s history, customer base, and the variety of workloads it has managed. Check customer references and ask for examples of how the provider supported individual customer needs.

Look for the widest variety of configuration options you can find so that you won’t be locked into a platform that either can’t scale to meet your needs or that delivers more compute power (and cost) than you require. For example, phoenixNAP offers more than 20 pre-configured bare-metal computing instances ranging from low-cost single CPU servers with 16GB of RAM to large dual-CPU instances with 768 GB of RAM, 4 TB of NVMe storage, and 50 Gbps of network capacity. New options are being added all the time.

Some cloud providers offer an alternative to bare- metal they call “dedicated servers.” While these are similar in some ways there are significant differences in areas like provisioning, billing, and customization. For example, server provisioning is fully automated in a bare-metal cloud but only partly orchestrated on a dedicated server, a difference that can translate into hours or days of wait time. A bare-metal cloud environment also offers more flexible billing and pre-configured instances for rapid deployment.

Download the full report, “Bare-Metal Computing Leads a Changing Cloud Landscape,” courtesy of phoenixNAP and Intel to learn more about bare-metal cloud computing. In our final article, we’ll discuss why processors matter when it comes to bare-metal computing. Catch up on previous articles in the series here and here

About the Author

Paul Gillin

Paul Gillin is a speaker, writer and technology journalist who has written five books and more than 400 articles on the topic of social media and digital marketing. A technology journalist for 25 years, he has served as Enterprise Editor of the tech news site SiliconAngle since 2014. He was founding editor-in-chief of B2B technology publisher TechTarget, and served as editor-in-chief and executive editor of the technology weekly Computerworld.

Sponsored Recommendations

The AI Disruption: Challenges and Guidance for Data Center Design

From large training clusters to small edge inference servers, AI is becoming a larger percentage of data center workloads. Learn more.

A better approach to boost data center capacity – Supply capacity agreements

Explore a transformative approach to data center capacity planning with insights on supply capacity agreements, addressing the impact of COVID-19, the AI race, and the evolving...

How Modernizing Aging Data Center Infrastructure Improves Sustainability

Explore the path to improved sustainability in data centers by modernizing aging infrastructure, uncovering challenges, three effective approaches, and specific examples outlined...

How Modern DCIM Helps Multi-Tenant Colocation Data Centers Be More Competitive

Discover the transformative impact of modern DCIM software on multi-tenant colocation data centers, enhancing competitiveness through improved resiliency, security, environmental...

Shutterstock, courtesy of Stream Data Centers

Why Liquid Cooling is More Sustainable

Stuart Lawrence of Stream Data Centers explains that because liquid is a more efficient means of heat transfer than air, liquid cooling will be necessary to support the next generation...

White Papers

Mgk Dcf Wp Cover2 2023 01 09 10 34 33

Data Center Microgrids: Planning for Your Microgrid

Jan. 9, 2023
The energy grid is increasingly vulnerable to outages thanks to aging infrastructure and the growing impact of climate change. Traditionally, data centers have turned to uninterruptible...