DigitalOcean is a developer-friendly Infrastructure-as-a-Service (IaaS) platform. With Cloudways, you can build and deploy high-performance, scalable applications on DigitalOcean with simplicity and freedom. By combining the Cloudways Managed Hosting Platform with DigitalOcean’s robust infrastructure, you get unmatched performance, reliability, and flexibility—backed by 24/7 support that works as your extended team.
Table of Contents
DigitalOcean Servers Family
To deliver the best possible hosting experience, DigitalOcean’s server lineup is divided into three main categories:
Basic Series
General Purpose Series
CPU-Optimized Series
Each family is designed for specific workload types. Choosing the right server for your workload is vital to your application’s success. Let’s take a closer look at these new instances to see what they offer.
1. Basic Series
Basic servers are perfect for everyday workloads and applications with moderate CPU usage. They leverage shared CPUs and provide a cost-effective solution for simpler projects, test environments, personal blogs, or small-scale websites. These servers are backed by SSD storage for the standard CPU type & nvme for the premium CPU type for faster data access than traditional disks.
Server Lineup
| Processor | Standard CPU Availability | Premium CPU Availability |
1GB | 1 vCPU | Yes | Yes |
2GB | 1 vCPU | Yes | Yes |
2GB | 2 vCPU | Yes | Yes |
4GB | 2 vCPU | Yes | Yes |
8GB | 4 vCPU | Yes | Yes |
16GB | 8 vCPU | Yes | Yes |
32GB | 8 vCPU | Yes | Yes |
48GB | 12 vCPU | Yes | No |
64GB | 16 vCPU | Yes | Yes |
96GB | 20 vCPU | Yes | No |
128GB | 24 vCPU | Yes | No |
192GB | 32 vCPU | Yes | No |
2. General Purpose
General Purpose servers are the next tier up from Basic and come with dedicated CPUs alongside a balanced memory-to-CPU ratio (4 GB of RAM per vCPU). These servers are designed for resource-intensive workloads, making them suitable for larger websites, e-commerce applications, and any project that needs both ample processing power and memory. They also leverage SSD storage for standard CPU type & nvme for premium CPU type to ensure quick data access and smooth performance for growing applications.
Server Lineup
| Processor | Standard CPU Availability | Premium CPU Availability |
8GB | 2 vCPU | Yes | Yes |
16GB | 4 vCPU | Yes | Yes |
32GB | 8 vCPU | Yes | Yes |
64GB | 16 vCPU | Yes | Yes |
128GB | 32 vCPU | Yes | Yes |
160GB | 40 vCPU | Yes | No |
192GB | 48 vCPU | No | Yes |
240GB | 60 vCPU | No | Yes |
3. CPU-Optimized
CPU-optimized servers are tailored for applications with higher processing demands. They feature dedicated CPUs and a 2:1 memory-to-CPU ratio (2 GB of RAM per vCPU), making them ideal for compute-intensive tasks such as large databases, complex web applications, and high-volume e-commerce. These servers deliver consistent performance and efficient processing power with dedicated CPU resources. Like other DigitalOcean servers, they come with SSD storage for the standard CPU type & nvme for the premium CPU type to ensure quick data access for resource-heavy workloads.
Server Lineup
| Processor | Standard CPU Availability | Premium CPU Availability |
4GB | 2 vCPU | Yes | No |
8GB | 4 vCPU | Yes | Yes |
16GB | 8 vCPU | Yes | Yes |
32GB | 16 vCPU | Yes | Yes |
64GB | 32 vCPU | Yes | Yes |
96GB | 48 vCPU | Yes | Yes |
120GB | 60 vCPU | No | Yes |
Benchmarks:
We have benchmarked DO instances from each family to compare the performance & webpage speed.
Note:
Our benchmarking focused on CPU-intensive testing, which may not fully reflect the performance characteristics of General Purpose instances. General Purpose instances are designed to offer a balanced allocation of CPU and memory resources, making them ideal for applications that require both processing power and memory bandwidth.
Example Use Cases:
Applications like:
Magento 2: Tends to consume more RAM due to the use of services like Elasticsearch, Redis, and RabbitMQ.
Laravel: Depending on the setup and dependencies, it may benefit from balanced CPU and memory provisioning.
For such workloads, General Purpose instances are often a more appropriate choice, ensuring consistent performance across both compute and memory-intensive operations.
Setup:
We used k6-wordpress-benchmarks methodology to setup the sites & the testing environment. The load storm test in k6 is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.
Performance Benchmark:
Configuration:
WordPress App with k6 loadouts.
Varnish Disabled.
Redis - Object Caching Disabled.
Caching Plugins removed.
Test Instances:
Basic Series - 8vCPU 16GB (Premium)
General Purpose Series - 8vCPU 16GB (Premium)
CPU-Optimized Series - 8vCPU 16GB (Premium)
Load Storm (k6) Benchmark Sequence:
Home Page: Initial request simulating a real visit.
Login Page: Navigates to
wp-login, submitting stored credentials.Sitemap Pages: Iterates through each URL found in the sitemap.
Asset Loading: On each page,
findNewAssets()retrieves CSS, JS, and images to replicate realistic resource loading while avoiding duplicates already fetched this session.
Note: K6 ramps from 1 to 1,000 users over 20 minutes, then hold peak load for 10 minutes.
Metrics:
The key metrics are:
Total Requests - Number of requests k6 made
p95 - The 95th percentile response time
Page Cum Avg - Cumulative average response time for Pages (html)
Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)
Instance Type | Total Requests | Avg P95 (ms) | Page Cum Avg (ms) | Asset Cum Avg (ms) | Login Cum Avg (ms) | WP-Login Avg Time (ms) |
Basic Premium | 394437 | 4014 | 2378 | 486 | 2156 | 2000 |
General Purpose Premium | 272014 | 10486 | 6524 | 1071 | 6194 | 6000 |
CPU Optimized Premium | 464078 | 1638 | 954 | 298 | 818 | 651 |
Virtual Users: 1,000, Duration: 30 minutes, Cache Status: No-Cache
Discussion
The three instance types show clear performance differences over a 30-minute, no-cache test with 1,000 virtual users:
CPU Optimized Premium stands out for both highest total requests (464,078) and the lowest latency figures across all metrics (e.g., Avg P95 of 1,638 ms). Its optimized CPU allocation handles concurrency very efficiently, leading to faster page loads and login times compared to the other tiers.
Basic Premium sits in the middle with 394,437 total requests and moderate response times (e.g., Avg P95 around 4,014 ms). While it can handle a fairly large volume, latency is higher than the CPU Optimized instance.
General Purpose Premium delivers the fewest total requests (272,014) and the highest response times (Avg P95 of 10,486 ms). Its broader, more balanced resource allocation struggles with high concurrency and large request volumes, leading to slower page and login responses.
Webpage speed Benchmark
WebPageTest fully loads the homepage and records how long it takes to load different aspects of the homepage. Result are measured in seconds.
Metrics:
The key metrics are:
Time to First Byte - When did the content start downloading?
First Contentful Paint - How soon did text and images start to appear?
Largest Contentful Paint - When did the largest visible content finish loading?
Page Speed Benchmark:
Configuration:
WordPress App with k6 loadouts.
Server Caching Varnish Disabled.
Redis - Object Caching Disabled.
Caching Plugins removed.
DataCenter: London
PageSpeed Location: London
Instance Type | Time to First Byte | First Contentful Paint | Largest Contentful Paint |
Basic Standard | 0.301 | 0.531 | 0.599 |
Basic Premium | 0.312 | 0.607 | 0.664 |
General Purpose Standard | 0.252 | 0.545 | 0.612 |
General Purpose Premium | 0.235 | 0.454 | 0.519 |
CPU Optimized Standard | 0.249 | 0.524 | 0.594 |
CPU Optimized Premium | 0.248 | 0.523 | 0.588 |
Discussion:
These results compare how quickly each instance delivers and renders content under a light-load scenario:
Time to First Byte (TTFB) measures how quickly the server responds. Here, General Purpose Premium is the fastest (0.235s), and Basic Premium is the slowest (0.312s).
First Contentful Paint (FCP) gauges when the first visible element appears. Again, General Purpose Premium excels (0.454s), indicating the page begins to render sooner for visitors.
Largest Contentful Paint (LCP) measures when the main content is fully visible. General Purpose Premium also leads (0.519s), while Basic Premium trails (0.664s).
Launching a New Application on a DigitalOcean Server
Step #1 — Accessing the Cloudways Platform
Log in to your Cloudways Platform using your credentials.
From the side menu bar, open My Servers.
Click Add Server.
Step #2 — Launching a New Server
Select the application you want to deploy.
Assign a name to your application.
Provide a name for your server.
Specify your project (optional).
Select DigitalOcean as your infrastructure provider.
Choose your preferred server type
Choose your preferred server size.
Choose your preferred CPU type
Important:
For the lowest latency, select the data center nearest to your primary audience.
Pick your server location from the list.
Finally, click Launch Now.
That’s it! Your new DigitalOcean server will be ready in a few minutes.
Tip:
If you’re new to Cloudways and have an existing website, you can request a managed migration from our support team, or use our Free Cloudways Migrator Plugin for WordPress sites.
Once your server and application(s) are created, make sure to follow the Take Site Live checklist to get your website online.
Upgrading an Existing Server to a New DigitalOcean Server Family
If you already use DigitalOcean, you can seamlessly upgrade to a higher-tier DO server type using the Vertical Scaling feature. If you’re on another cloud provider and wish to switch, you can clone your server over to DigitalOcean.
That’s it! We hope this guide helps. If you need further assistance, feel free to search the Cloudways Support Center or contact our support team via live chat (Need a Hand > Send us a Message). Alternatively, you can open a support ticket.