QUM Performance

Comparing the performance across multiple providers

Last updated 2 months ago

QUM Performance

Comparing the performance across multiple providers

We benchmarked VMs across different providers namely aws, gcp, azure and digital ocean to compare how Qumulus VMs perform and where dose it stand! and was used on all VMs running Ubuntu 24.04

VMs were created on all of these platform using 2 vCPUs and 8 GB RAM with instance types and tentative monthly cost shown in the table below (UK region). These costs on other platform would vary based on the storage and network utilization.

With Qumulus, its slightly different. Qumulus does not sell a VM of 2 vCPUs and 8 GB RAM. Instead a Compute Unit of 8 vCPUs, 20 GB RAM and 200 GB Storage is sold for $ 85/month. So after creating a VM of 2 vCPUs and 8 GB RAM, there is still more than double capacity available for additional use. We may assume cost of running this VM as $ 30/month.

Instance Types and pricing

Provider

Instance Type

vCPUs

RAM

Monthly Cost

Qumulus Cloud

m1.medium

Digital Ocean

Premium AMD

AWS

m5a.large

Azure

Standard D2s v4

GCP

e2-standard-2

Results

These results will surprise you. Multicore performance of Qumulus is 2-to-4 times better than other providers. Qumulus is at least 2 times more efficient than the 2nd most efficient provider in Performance to Cost ratio.

Geekbench6 Results

Sysbench CPU Results

Sysbench Memory Results

Performance to Cost Ratio

How to Run benchmark?

Running Geekbench6 on Ubuntu

mkdir downloads
cd ~/downloads/
wget https://cdn.geekbench.com/Geekbench-6.2.1-Linux.tar.gz
tar xf Geekbench-6.2.1-Linux.tar.gz
cd Geekbench-6.2.1-Linux
./geekbench6

Running Sysbench on Ubuntu

sudo apt update && sudo apt install -y sysbench
sysbench cpu --cpu-max-prime=20000 run
sysbench memory --memory-total-size=10G run

QUM Performance