VPS provider comparison table<\/a> to see how different hosts stack up on performance specifications.<\/p>\n\n\n\n1. CPU Benchmarking<\/h2>\n\n\n\n
Two vCPUs from one provider can deliver 2-3x the performance of two vCPUs from another, depending on the CPU generation and whether the cores are pinned or shared. Run these tests to measure your actual CPU power.<\/p>\n\n\n\n
Using sysbench<\/h3>\n\n\n\n
sysbench is a lightweight benchmarking tool available in most package managers. It calculates prime numbers to measure CPU computational throughput.<\/p>\n\n\n\n
# Install sysbench\nsudo apt install sysbench -y\n\n# Single-core performance test\nsysbench cpu run --threads=1 --cpu-max-prime=20000\n\n# Multi-core (all available cores)\nsysbench cpu run --threads=$(nproc) --cpu-max-prime=20000<\/code><\/pre>\n\n\n\nInterpreting results:<\/strong> The “events per second” metric is the number of prime calculations completed. Compare this across providers \u2014 a difference of 30% or more indicates a significant CPU generation gap. A modern AMD EPYC 9654 (Genoa) core should score roughly 2x higher than an older Intel Xeon E5-2690 v2 core.<\/p>\n\n\n\nUsing openssl Speed Test<\/h3>\n\n\n\n
openssl provides a quick CPU benchmark without installing extra packages:<\/p>\n\n\n\n
# Multi-core AES encryption benchmark\nopenssl speed -elapsed -multi $(nproc) aes-256-cbc<\/code><\/pre>\n\n\n\nThis is particularly relevant for VPS instances used for TLS termination (web servers, VPNs) where AES-NI hardware acceleration makes a significant difference.<\/p>\n\n\n\n
2. Disk I\/O Benchmarking<\/h2>\n\n\n\n
Storage performance is the most common bottleneck in VPS hosting. A provider advertising “SSD storage” may be using SATA SSDs (500 MB\/s) or NVMe (3,500+ MB\/s). Always test with fio.<\/p>\n\n\n\n
Sequential Read\/Write<\/h3>\n\n\n\n
Sequential benchmarks measure throughput for large file operations \u2014 useful for file servers, backups, and media streaming.<\/p>\n\n\n\n
# Install fio\nsudo apt install fio -y\n\n# Sequential read (1M block size)\nfio --name=seq-read --rw=read --bs=1M --size=1G --runtime=30 --direct=1\n\n# Sequential write\nfio --name=seq-write --rw=write --bs=1M --size=1G --runtime=30 --direct=1<\/code><\/pre>\n\n\n\nRandom 4K IOPS (Database Workload)<\/h3>\n\n\n\n
Random 4K reads simulate database lookups (MySQL, PostgreSQL). This is the most important VPS benchmark for transactional workloads.<\/p>\n\n\n\n
# Random read IOPS (4K block, 32 depth)\nfio --name=rand-read --rw=randread --bs=4k --size=1G --runtime=60 --iodepth=32 --direct=1\n\n# Random write IOPS\nfio --name=rand-write --rw=randwrite --bs=4k --size=1G --runtime=60 --iodepth=32 --direct=1<\/code><\/pre>\n\n\n\nTargets:<\/strong> NVMe storage should achieve 300,000+ IOPS for 4K random reads. SATA SSDs typically deliver 50,000-90,000 IOPS. If you see fewer than 10,000 IOPS, you may be on shared HDD storage.<\/p>\n\n\n\n3. Network Speed Benchmarking<\/h2>\n\n\n\n
Network throughput on a VPS is limited by the virtual port speed, the provider’s upstream bandwidth, and the quality of the hypervisor’s virtual NIC driver.<\/p>\n\n\n\n
Using iperf3<\/h3>\n\n\n\n
iperf3 measures TCP and UDP throughput between two endpoints. You need a test server on the other end \u2014 many public iperf3 servers are available.<\/p>\n\n\n\n
# Install iperf3\nsudo apt install iperf3 -y\n\n# Test against a public iperf3 server (NYC)\niperf3 -c iperf.he.net -t 30\n\n# Reverse mode (download test)\niperf3 -c iperf.he.net -t 30 -R<\/code><\/pre>\n\n\n\nQuick Download Test<\/h3>\n\n\n\n
If you don’t have access to an iperf3 server, use a direct download speed test:<\/p>\n\n\n\n
# Download a 100MB test file and measure throughput\ncurl -o \/dev\/null -s -w \"Speed: %{speed_download} bytes\/sec\n\" https:\/\/speedtest-nyc1.digitalocean.com\/100mb.test<\/code><\/pre>\n\n\n\n4. Latency and Network Quality Testing<\/h2>\n\n\n\n
Latency affects application responsiveness, especially for database queries, API calls, and real-time services.<\/p>\n\n\n\n