High Performance and Truly EPYC: Vultr’s New VMs Are Ready for Anything
Mar 15, 2022

Today we’re launching a completely new Optimized Cloud Compute product line, along with a refreshed Cloud Compute lineup of virtual machines. Overall, we’re introducing a wide range of shapes and sizes of High Performance VMs, all with 100% NVMe SSD storage, and including instances powered by 3rd Generation AMD EPYC™ processors.

This is kind of a big deal for us, and hopefully for you too. Now more than ever, you might want to use Vultr for everything ranging from mission-critical enterprise needs to your personal projects.

Compute for Any Use Case
We have made a number of updates to our platform, but before we dive into the details, here is a bird’s-eye view.

You can think of our Compute platform as having three discrete product lines:

1. Optimized Cloud Compute – Starting at $28 per month, these VMs are easy to use and elastically scalable. Because they run atop dedicated vCPUs, your workloads run with fast consistent performance, with no noisy neighbors. This quality makes these instances the best choice for many business applications, such as high traffic websites, transcoding, CI/CD, etc.

2. Cloud Compute – Starting at just $5 per month, these VMs are highly affordable because they run on shared vCPUs. For many bursty applications (e.g., low traffic websites, game servers, VPNs, etc.), Cloud Compute instances deliver great performance at a remarkably low cost.

3. Bare Metal – Starting at $120 per month, these servers provide dedicated physical CPUs and direct access to your server's underlying hardware, thereby delivering maximum performance without any virtualization overhead. Bare Metal is as easy to provision and manage as our VMs, and for particularly intense workloads it can deliver the best price-to-performance.

You’ll now see these product categories when you deploy a server within our control panel:




Powered by 3rd Generation AMD EPYC CPUs and 100% NVMe SSD Storage
All Optimized Cloud Compute instances, and select High Performance Cloud Cloud instances, are powered by 3rd Gen AMD EPYC CPUs.

This is particularly obvious when you provision a Cloud Compute instance in the control panel. Within that section, you’ll see new options for CPU & Storage Technology.



This is the first time we’ve offered AMD processor-powered virtual machines, and given their exceptional performance, these VMs are now our default, and recommended option for most users.

AMD has pushed the performance and technology capabilities of the EPYC family of processors over the last few years. 3rd generation EPYC processors, which power our new VMs, continue that trajectory, providing up to 64 cores in a single CPU. The particular 3rd Gen EPYC processors that underpin our new VMs are powered by AMD’s Zen 3 microarchitecture, which delivers fantastic performance per core along with a maximum frequency of 3.675GHz.

We obviously still love Intel too, and if Intel-powered instances are the right choice for your needs, we have a wide range of plans for you!

Optimized Cloud Compute is Powered by Dedicated vCPUs, Built with 100% NVMe SSD Storage, and Available in a Variety of Server Types
Today’s launch includes the introduction of a new product line called Optimized Cloud Compute.

Unlike our typical Cloud Compute instances that run on shared vCPUs, Optimized Cloud Compute instances run on vCPUs that are dedicated to your workload, so you never have to worry about noisy neighbors.

Optimized Cloud Compute instances are available in a variety of server types, with varying levels of CPU, RAM, and storage to best support a broad range of workloads.



Optimized Cloud Compute instances are designed to be a more affordable alternative to solutions like AWS EC2, in order to help users transition from the Big Tech clouds to our independent platform.

Here’s how to think about which server type is best for you:

General Purpose – Often a good starting point, these VMs provide a typical balance of CPU, RAM, and NVMe SSD storage. We recommend General Purpose instances for many web and application servers, e-commerce, game servers, video and audio streaming, API serving, and relational databases.

CPU Optimized – For compute bound applications, these VMs provide proportionally more CPU than they do RAM and NVMe SSD storage. We recommend CPU Optimized instances for workloads like video encoding, batch processing, CI/CD, high performance computing (HPC), ad serving, and analytics processing.

Memory Optimized – For memory bound applications, these VMs provide proportionally more RAM than they do CPU and NVMe SSD storage. We recommend Memory Optimized instances for open source databases (e.g., MySQL), in-memory databases and caches (e.g., Memcached), and real-time analytics.

Storage Optimized – These VMs provide especially generous proportions of NVMe SSD storage accompanying a typical balance of CPU and RAM. We recommend Storage Optimized instances for large non-relational databases (e.g., Cassandra and MongoDB), and high frequency online transaction processing (OLTP).

Get Started Today – All Around the Globe
Vultr is available in 24 locations across five continents:

North America: Atlanta, Chicago, Dallas, Honolulu, L.A., Mexico City, Miami, New Jersey, Seattle, Silicon Valley, Toronto
South America: São Paulo
Europe: Amsterdam, Frankfurt, London, Madrid, Paris, Stockholm, Warsaw
Asia: Seoul, Singapore, Tokyo
Australia: Melbourne, Sydney

As always, you can provision new instances via our control panel, API, or CLI.

Businesses interested in our new compute offerings should contact sales to learn more. Managed service providers and resellers interested in bringing the Vultr platform to their customers are welcome to join our partner program.