Skip to main content

Hetzner CPX32

Falkenstein DC Park 1

Benchmark complete
Finished Oct 17, 2026
Geekbench 7 single-core
1,670
Faster than 87% of 4 vCPU plans
Geekbench 7 multi-core
5,910
Faster than 90% of 4 vCPU plans
Disk 4K IOPS
47,220
fio random read and write, 50/50
Monthly price at test time
$41.99
141 multi-core points per dollar
Same price today

Server

  • Provider
    Hetzner
  • Plan
    CPX32
  • Location
    Falkenstein DC Park 1
  • vCPU
    4
  • RAM
    8 GB
  • Storage
    160 GB NVMe
  • Network
    Unknown
  • OS image
    ubuntu-24.04
  • Trigger
    Scheduled
  • Queued at
    Oct 17, 2026, 21:24:00 UTC
Visit Hetzner

CPU

Version Single-core Multi-core
Geekbench 7 1,670 5,910 Geekbench Browser

Scores from different Geekbench versions aren't comparable with each other. Charts only use Geekbench 7.

Geekbench 7 workloads

Workload Single-core Multi-core
File Compression
1,703
41st of 241
4,951
21st of 166
Navigation
1,731
26th of 241
Single-core only
HTML5 Browser
1,640
38th of 241
Single-core only
PDF Viewer
1,742
40th of 241
Single-core only
Photo Library
1,748
20th of 241
6,416
21st of 166
Clang
1,471
39th of 241
6,054
24th of 166
Text Processing
1,524
40th of 241
5,612
22nd of 166
Asset Compression
1,719
39th of 241
6,788
27th of 166
HDR
1,741
40th of 241
5,022
20th of 166
Photo Editor
1,989
33rd of 241
6,626
13th of 166
Ray Tracer
1,558
46th of 241
6,107
28th of 166
Structure from Motion
1,717
54th of 241
Single-core only
Game Physics
1,612
36th of 241
Single-core only
Video Encoder
1,559
45th of 241
Single-core only
Audio Encoder
1,624
48th of 241
Single-core only
Video Player
1,702
52nd of 241
Single-core only

Rank among the 241 runs with Geekbench 7 workload scores. Bars are relative to the best run. Geekbench doesn't run every workload multi-core, so some rows only have a single-core score.

Memory, TLS and scheduler

Memory bandwidth (STREAM)

Test Result Runs Spread
Copy higher is better 88.2 GB/s 3 17.5%
Scale higher is better 71.5 GB/s 3 18.2%
Add higher is better 69.7 GB/s 3 17.1%
Triad higher is better 70.6 GB/s 3 17.4%

Average of 3 runs. Spread is the standard deviation as a share of the average.

TLS and crypto (OpenSSL)

Test Result Runs Spread
RSA 4096 signs higher is better 2,608/s 3 0.612%
RSA 4096 verifies higher is better 63,498/s 3 1.06%
SHA-256 higher is better 3.91 GB/s 3 0.467%
AES-256-GCM higher is better 39 GB/s 3 0.105%
ChaCha20-Poly1305 higher is better 14.7 GB/s 3 0.805%

Average of 3 runs. Spread is the standard deviation as a share of the average.

Scheduler latency (schbench)

Test Result Runs Spread
Request latency, median lower is better 4.12 ms 3 0.225%
Requests per second, 90th percentile higher is better 880/s 3 1.14%
Request latency, 99.9th percentile lower is better 13.3 ms 3 5.24%
Requests per second, average higher is better 844/s 3 0.312%

Average of 3 runs. Spread is the standard deviation as a share of the average.

Disk

Block size Read Write Total
4K 92.1 MB/s (23,579 IOPS) 92.3 MB/s (23,641 IOPS) 184.5 MB/s (47,220 IOPS)
64K 1.25 GB/s (20,528 IOPS) 1.26 GB/s (20,636 IOPS) 2.51 GB/s (41,164 IOPS)
512K 1.29 GB/s (2,635 IOPS) 1.36 GB/s (2,775 IOPS) 2.64 GB/s (5,410 IOPS)
1M 1.51 GB/s (1,548 IOPS) 1.61 GB/s (1,651 IOPS) 3.13 GB/s (3,199 IOPS)

fio random read and write, 50/50, run by YABS against the boot disk.

Network

IPv4 transfer speed

Loading...
Target Protocol Send Receive
Clouvider, London, UK (10G) IPv4 7.41 Gbps 7.24 Gbps
Eranium, Amsterdam, NL (100G) IPv4 5.8 Gbps 4.16 Gbps
Leaseweb, Singapore, SG (10G) IPv4 1 Gbps 1.42 Gbps
Clouvider, Los Angeles, CA, US (10G) IPv4 1 Gbps 928 Mbps
Leaseweb, NYC, NY, US (10G) IPv4 2.3 Gbps 2.62 Gbps

Public iperf3 servers vary in capacity and availability. Compare a target only with itself over time.

Datastores

PostgreSQL

fsync speed (pg_test_fsync)

Method Operations per second Time per operation
One 8 kB write
open_datasync 2,551 0.392 ms
fdatasync 2,413 0.414 ms
fsync 829 1.21 ms
fsync_writethrough Not supported here n/a
open_sync 877 1.14 ms
Two 8 kB writes
open_datasync 1,267 0.789 ms
fdatasync 2,237 0.447 ms
fsync 855 1.17 ms
fsync_writethrough Not supported here n/a
open_sync 435 2.3 ms
open_sync at different write sizes
1 * 16kB open_sync write 964 1.04 ms
2 * 8kB open_sync writes 459 2.18 ms
4 * 4kB open_sync writes 252 3.98 ms
8 * 2kB open_sync writes 121 8.26 ms
16 * 1kB open_sync writes 55.9 17.9 ms
fsync on a non-write descriptor
write, fsync, close 914 1.09 ms
write, close, fsync 929 1.08 ms
Non-synced 8 kB writes
write 612,871 0.002 ms

pg_test_fsync against the Docker volume Postgres keeps its data on. fdatasync is Linux's default wal_sync_method.

pgbench at scale 10

Workload Clients Transactions/s Median p95 p99 p99.9 Failed
Read/write (TPC-B like) 1 754 1.32 ms 1.7 ms 2.21 ms 3.64 ms 0
Read/write (TPC-B like) 8 4,903 1.43 ms 3.05 ms 4.72 ms 8.3 ms 0
Read/write (TPC-B like) 32 5,700 4.06 ms 14.8 ms 26.7 ms 47.7 ms 0
Read only 1 15,549 0.058 ms 0.11 ms 0.155 ms 0.248 ms 0
Read only 8 119,943 0.057 ms 0.123 ms 0.185 ms 0.314 ms 0
Read only 32 132,713 0.198 ms 0.326 ms 0.448 ms 0.783 ms 0

Scale 10 (1,000,000 accounts), loaded in 1.24 s. 60 s per run, prepared statements. Latency comes from sampled transaction logs, and p99.9 is left out when fewer than 1,000 were sampled.

pgbench at scale 100

Workload Clients Transactions/s Median p95 p99 p99.9 Failed
Read/write (TPC-B like) 1 738 1.34 ms 1.73 ms 3.02 ms 8.07 ms 0
Read/write (TPC-B like) 8 5,276 1.34 ms 2.49 ms 6.32 ms 11.5 ms 0
Read/write (TPC-B like) 32 7,391 3.71 ms 8.19 ms 13.3 ms 25.4 ms 0
Read only 1 12,493 0.071 ms 0.134 ms 0.189 ms 0.299 ms 0
Read only 8 92,713 0.074 ms 0.165 ms 0.246 ms 0.459 ms 0
Read only 32 96,069 0.286 ms 0.499 ms 0.701 ms 1.21 ms 0

Scale 100 (10,000,000 accounts), loaded in 11.4 s. 60 s per run, prepared statements. Latency comes from sampled transaction logs, and p99.9 is left out when fewer than 1,000 were sampled.

MySQL

sysbench OLTP

Workload Threads Transactions/s Queries/s Median p95 p99 Errors
Read/write 1 169 3,389 5.67 ms 7.84 ms 13.7 ms 0
Read/write 8 1,132 22,644 6.43 ms 11.9 ms 19.3 ms 0
Read/write 32 1,784 35,689 17 ms 30.8 ms 41.1 ms 0
Read only 1 641 10,250 1.37 ms 2.57 ms 3.08 ms 0
Read only 8 5,037 80,590 1.52 ms 2.22 ms 3.25 ms 0
Read only 32 4,658 74,521 6.55 ms 9.39 ms 13.2 ms 0

10 tables of 100,000 rows, loaded in 5.3 s. 60 s per run, one connection per thread.

Redis

memtier

Mix Operations/s Hits/s Median p99 p99.9
1:10 set:get 73,063 66,420 0.831 ms 1.88 ms 3.46 ms
1:1 set:get 75,050 37,525 0.807 ms 1.89 ms 3.44 ms

1,000,000 keys of 32 bytes. 64 connections over 4 threads, 60 s per ratio.

Web frameworks

Framework Fortunes req/s Single query req/s JSON req/s Fortunes p99 at 256 connections
Next.js nextjs 1,576 5,736 8,388 282 ms
Express express-postgresjs 16,856 23,468 74,574 35.1 ms
Fastify fastify-postgres 17,829 24,484 78,522 29.3 ms
NestJS nestjs 11,888 13,798 54,533 43.3 ms
Nuxt nuxt 15,713 24,454 75,747 39.6 ms
FastAPI fastapi 9,259 11,243 49,253 181 ms
Django django-postgresql 4,407 5,065 13,758 176 ms
Rails rails 7,044 9,207 43,874 61.6 ms
Laravel laravel 1,799 1,979 3,970 164 ms
Symfony symfony 2,318 2,670 7,789 126 ms

Peak requests per second across the 64 and 256 connection levels, measured with wrk. Each test has to pass the TechEmpower response checks first, or it's marked invalid.

Web apps

WooCommerce

Page mix

Step Requests/s Median p95 p99 Error rate
8 virtual users
Home page 5.84 275 ms 361 ms 407 ms 0%
Shop page 5.84 363 ms 483 ms 523 ms 0%
Category page 5.84 232 ms 387 ms 437 ms 0%
Product page 5.84 256 ms 343 ms 367 ms 0%
Product search 5.84 223 ms 314 ms 341 ms 0%
32 virtual users
Home page 5.59 1.1 s 1.28 s 1.34 s 0%
Shop page 5.59 1.41 s 1.63 s 1.73 s 0%
Category page 5.59 1.11 s 1.43 s 1.56 s 0%
Product page 5.59 1.07 s 1.21 s 1.27 s 0%
Product search 5.59 962 ms 1.1 s 1.14 s 0%

Cart and checkout

Step Requests/s Median p95 p99 Error rate
8 virtual users
Add to cart 11.3 278 ms 381 ms 418 ms 0%
Cart 11.3 240 ms 319 ms 358 ms 0%
Checkout 11.3 189 ms 261 ms 296 ms 0%
32 virtual users
Add to cart 11 1.07 s 1.28 s 1.34 s 0%
Cart 11 971 ms 1.16 s 1.24 s 0%
Checkout 11 824 ms 987 ms 1.03 s 0%

k6 with a fixed number of virtual users, 61 to 63 s per level. The error rate counts responses that weren't a 200.

Discourse

Signed in

Step Requests/s Median p95 p99 Error rate
1 virtual user
Topic page 1.99 237 ms 279 ms 466 ms 0%
Topic JSON 0.994 114 ms 140 ms 149 ms 0%
Home page 1.01 195 ms 223 ms 303 ms 0%
Categories page 0.497 231 ms 278 ms 358 ms 0%
User activity page 0.497 155 ms 210 ms 258 ms 0%
4 virtual users
Topic page 9.68 182 ms 245 ms 424 ms 0%
Topic JSON 4.86 92.2 ms 124 ms 201 ms 0%
Home page 4.86 163 ms 216 ms 371 ms 0%
Categories page 2.43 189 ms 255 ms 457 ms 0%
User activity page 2.41 115 ms 232 ms 317 ms 0%
16 virtual users
Topic page 9.39 707 ms 940 ms 1.21 s 0%
Topic JSON 4.73 520 ms 653 ms 800 ms 0%
Home page 4.75 669 ms 886 ms 1.13 s 0%
Categories page 2.37 732 ms 933 ms 1.14 s 0%
User activity page 2.33 572 ms 821 ms 934 ms 0%

Anonymous

Step Requests/s Median p95 p99 Error rate
16 virtual users
Topic page 205 19.5 ms 38.6 ms 55.8 ms 0%
Topic JSON 103 17.8 ms 36.1 ms 55.2 ms 0%
Home page 103 19.8 ms 38.3 ms 54.3 ms 0%
Categories page 51.3 18.8 ms 36.9 ms 59.5 ms 0%
User activity page 51.3 114 ms 153 ms 257 ms 0%

k6 with a fixed number of virtual users, 60 to 61 s per level, after a 21 s warmup. The error rate counts responses that weren't a 200. Discourse ran with 8 app workers and 2048 MB of Postgres shared buffers, what its installer picks for a 8 GB box. Its installer adds no swap on a box this big. The container peaked at 2,714 MB.

Ghost

Readers

Step Requests/s Median p95 p99 Error rate
1 virtual user
Home page 9.31 27.6 ms 44.4 ms 51.7 ms 0%
Post page 23.2 14.8 ms 26.6 ms 30.5 ms 0%
Tag page 4.65 18.9 ms 32 ms 34.7 ms 0%
Author page 4.65 18.6 ms 30.7 ms 33.7 ms 0%
RSS feed 4.65 22.3 ms 37.6 ms 41.9 ms 0%
4 virtual users
Home page 14 77.5 ms 106 ms 119 ms 0%
Post page 35.1 48.4 ms 76.6 ms 90.4 ms 0%
Tag page 7.04 49.1 ms 72.6 ms 88.5 ms 0%
Author page 7.01 48.7 ms 70.8 ms 87.1 ms 0%
RSS feed 7.01 57.4 ms 85 ms 100 ms 0%
16 virtual users
Home page 13.9 292 ms 344 ms 365 ms 0%
Post page 34.7 243 ms 300 ms 324 ms 0%
Tag page 6.93 161 ms 216 ms 232 ms 0%
Author page 6.97 157 ms 205 ms 223 ms 0%
RSS feed 6.93 165 ms 218 ms 240 ms 0%

Staff in Ghost Admin

Step Requests/s Median p95 p99 Error rate
1 virtual user
Post list 4.3 35.3 ms 48.2 ms 52.5 ms 0%
Post editor 12.9 12.1 ms 19.9 ms 22.5 ms 0%
Saving a draft 12.9 35.4 ms 48.4 ms 54.7 ms 0%
Member list 4.32 39.9 ms 53.9 ms 64 ms 0%
4 virtual users
Post list 6.98 36.1 ms 52.2 ms 58.5 ms 0%
Post editor 21 16.4 ms 28.8 ms 33.1 ms 0%
Saving a draft 21.1 141 ms 180 ms 200 ms 0%
Member list 7.03 39.9 ms 58.3 ms 64.4 ms 0%

k6 with a fixed number of virtual users, 60 s per level, after a 20 s warmup. The error rate counts responses that weren't a 200. Ghost ran the way its ghost-docker setup runs it: one Node process, and MySQL with its stock 128 MB buffer pool. Memory peaked with Ghost at 540 MB and MySQL at 624 MB.

Wagtail

Visitors

Step Requests/s Median p95 p99 Error rate
1 virtual user
Home page 1.43 86.2 ms 107 ms 139 ms 0%
Blog index 1.43 309 ms 373 ms 399 ms 0%
Blog post 4.24 33.9 ms 47.5 ms 56.5 ms 0%
Bread page 4.26 35.9 ms 48.8 ms 53.6 ms 0%
Location page 1.43 35.3 ms 43 ms 44.1 ms 0%
Site search 1.41 53.6 ms 84.8 ms 109 ms 0%
4 virtual users
Home page 2.64 219 ms 394 ms 426 ms 0%
Blog index 2.66 658 ms 1.26 s 1.39 s 0%
Blog post 7.9 52.2 ms 114 ms 151 ms 0%
Bread page 7.92 55.5 ms 123 ms 153 ms 0%
Location page 2.67 56.5 ms 111 ms 135 ms 0%
Site search 2.62 77.8 ms 228 ms 355 ms 0%
16 virtual users
Home page 3.52 533 ms 704 ms 803 ms 0%
Blog index 3.55 1.43 s 1.71 s 1.8 s 0%
Blog post 10.5 296 ms 474 ms 525 ms 0%
Bread page 10.6 309 ms 472 ms 549 ms 0%
Location page 3.55 289 ms 457 ms 538 ms 0%
Site search 3.5 352 ms 507 ms 592 ms 0%

Editors in the Wagtail admin

Step Requests/s Median p95 p99 Error rate
1 virtual user
Admin dashboard 0.762 87.3 ms 112 ms 123 ms 0%
Page explorer 1.56 295 ms 379 ms 519 ms 0%
Page editor 2.3 146 ms 176 ms 279 ms 0%
Image library 0.762 62.5 ms 78 ms 443 ms 0%
Admin page search 0.762 42.3 ms 79.6 ms 83 ms 0%
4 virtual users
Admin dashboard 0.924 314 ms 893 ms 942 ms 0%
Page explorer 1.88 923 ms 1.92 s 2.77 s 0%
Page editor 2.85 437 ms 1.06 s 1.23 s 0%
Image library 0.957 234 ms 597 ms 791 ms 0%
Admin page search 0.94 121 ms 314 ms 527 ms 0%

k6 with a fixed number of virtual users, 60 to 61 s per level, after a 21 s warmup. The error rate counts responses that weren't a 200. Wagtail ran the way bakerydemo ships it: uWSGI with 2 workers of 4 threads each, whatever the box, and Postgres and Redis on stock settings. Memory peaked with Wagtail at 296 MB, Postgres at 151 MB, and Redis at 25 MB.

Advertised vs. measured

Advertised Measured
vCPU 4 4
CPU Shared vCPU 0% of busy CPU time stolen
RAM 8 GB 7.6 GiB
Storage 160 GB NVMe 150.1 GiB
Architecture x86_64 x86_64
Virtualization KVM KVM

Guests usually report slightly less RAM and disk than advertised because of kernel reservations and filesystem overhead.

Detected hardware

CPU model
AMD EPYC-Genoa Processor
CPU cores
4
CPU frequency
2400 MHz
AES-NI
Yes
VM-x/AMD-V
No
RAM reported
7.6 GiB
Swap reported
0 MiB
Disk reported
150.1 GiB
Virtualization reported
KVM
CPU steal
0% of busy CPU time stolen
Distro
Ubuntu 24.04.4 LTS
Kernel
6.8.0-138-generic
Architecture
x86_64
IPv4 reachable
Yes
IPv6 reachable
Yes

Advertised at test time

Plan
CPX32
Product line
regular_purpose
Location
Falkenstein DC Park 1
vCPU
4 shared vCPU
Architecture
x86_64
RAM
8 GB
Storage
160 GB NVMe
Virtualization
KVM
Network
Unknown
Transfer
21.99 TB
Port 25
On request
Monthly price at test time
$41.99
Hourly price at test time
$0.067
YABS version
v2026-09-20
Ran at
Oct 17, 2026, 21:24:49 UTC
Parser version
5

Run timeline

Queued
Oct 17, 2026, 21:24:00 UTC
First update from the server
Oct 17, 2026, 21:24:42 UTC
Completed
Oct 17, 2026, 23:00:27 UTC
Server deleted
Oct 17, 2026, 23:00:27 UTC
Cost of this run
$0.11
Watch party 0 watching

This party closed at . It's read only now.

Watch party

Loading the party