• We value your experience with Plesk during 2025
    Plesk strives to perform even better in 2026. To help us improve further, please answer a few questions about your experience with Plesk Obsidian 2025.
    Please take this short survey:

    https://survey.webpros.com/

Input Plesk can now be installed on Ubuntu 20 servers running on ARM-based platforms

We are a small, London-based hosting company without plans to rapidly grow. We don't aim to offer the cheapest hosting, because there are multi-million dollar American companies that wipe the floor on that. You probably have ten or a hundred times as many customers as we do. But we've been doing this since 1998, so we must be doing something right, albeit in a niche market. Our customers are agencies for the most part.

When considering the cost of providing hosting services, it's not all about the cost of the server, but over a 3 - 5 year lifespan, it's also about the total cost of ownership. As you'll be aware, Ampere CPUs are much more efficient than the Xeon or AMD CPUs of a similar vintage (2022 in this case). Not all datacentres charge for metered power, but an increasing number do. So even if your customers don't care about servers that use less kWH, your accountant/CFO probably does. And although it may seem that Ampere is an outlier technology, it's worth bearing in mind that Amazon, Google and Microsoft all offer virtual compute instances in their clouds, some using Ampere CPUs, and some using proprietary ARM hardware technology such as Graviton which the rest of us aren't allowed to purchase on the open market.
 
We are a small, London-based hosting company without plans to rapidly grow. We don't aim to offer the cheapest hosting, because there are multi-million dollar American companies that wipe the floor on that. You probably have ten or a hundred times as many customers as we do. But we've been doing this since 1998, so we must be doing something right, albeit in a niche market. Our customers are agencies for the most part.

When considering the cost of providing hosting services, it's not all about the cost of the server, but over a 3 - 5 year lifespan, it's also about the total cost of ownership. As you'll be aware, Ampere CPUs are much more efficient than the Xeon or AMD CPUs of a similar vintage (2022 in this case). Not all datacentres charge for metered power, but an increasing number do. So even if your customers don't care about servers that use less kWH, your accountant/CFO probably does. And although it may seem that Ampere is an outlier technology, it's worth bearing in mind that Amazon, Google and Microsoft all offer virtual compute instances in their clouds, some using Ampere CPUs, and some using proprietary ARM hardware technology such as Graviton which the rest of us aren't allowed to purchase on the open market.


You make a fair point about total cost of ownership, and I absolutely agree that power efficiency is becoming more relevant; especially in datacentres with metered power billing.

That said, in our segment of the market (primarily SMB and price-sensitive hosting customers), the upfront and predictable cost per VM or per hosting account still weighs heavier than long-term theoretical efficiency gains. Most clients don’t ask what CPU architecture we run on — they ask what it costs per month and whether it performs reliably.

From a technical perspective, ARM is interesting, and Ampere has clearly positioned itself well. But for us, the decision also involves ecosystem maturity and operational risk. x86 remains the lowest-friction option in terms of software compatibility, third-party tooling, backup agents, security tooling, and predictable behaviour under mixed workloads. Especially in environments where Plesk, extensions, Imunify, backup software, and other commercial components all need to work seamlessly.

If ARM adoption continues to grow and more vendors officially support aarch64 as a first-class citizen rather than a “we’ll see” platform, that lowers the barrier significantly. At that point it becomes less of a technical experiment and more of a strategic alternative.

For now, we optimise heavily on density, consolidation and efficiency within the x86 ecosystem. That gives us a very predictable cost model across 3–5 years, which matters when you operate at scale.

But I do appreciate pioneers pushing ARM forward — broader adoption only happens if someone actually deploys it in production and shares real-world results.
 
Back
Top