Skip to main content

Hetzner CCX23

Helsinki DC Park 1

Benchmark complete
Finished Oct 12, 2026
Geekbench 7 single-core
1,738
Faster than 100% of 4 vCPU plans
Geekbench 7 multi-core
4,369
Faster than 75% of 4 vCPU plans
Disk 4K IOPS
131,267
fio random read and write, 50/50
Monthly price at test time
$101.49
43 multi-core points per dollar
Same price today

Server

  • Provider
    Hetzner
  • Plan
    CCX23
  • Location
    Helsinki DC Park 1
  • vCPU
    4
  • RAM
    16 GB
  • Storage
    160 GB NVMe
  • Network
    Unknown
  • OS image
    ubuntu-24.04
  • Trigger
    Manual
  • Queued at
    Oct 12, 2026, 01:06:45 UTC
Visit Hetzner

Latest runs of this plan

Full history of this plan

CPU

Version Single-core Multi-core
Geekbench 7 1,738 4,369 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
1,698
25th of 106
4,380
16th of 90
Navigation
1,336
53rd of 106
Single-core only
HTML5 Browser
1,781
17th of 106
Single-core only
PDF Viewer
1,971
13th of 106
Single-core only
Photo Library
1,600
22nd of 106
4,304
22nd of 90
Clang
1,551
20th of 106
4,529
24th of 90
Text Processing
1,757
12th of 106
3,900
28th of 90
Asset Compression
1,802
15th of 106
4,791
27th of 90
HDR
1,843
15th of 106
3,835
21st of 90
Photo Editor
2,092
16th of 106
4,703
20th of 90
Ray Tracer
1,662
18th of 106
4,614
27th of 90
Structure from Motion
1,758
22nd of 106
Single-core only
Game Physics
1,608
22nd of 106
Single-core only
Video Encoder
1,771
17th of 106
Single-core only
Audio Encoder
1,785
17th of 106
Single-core only
Video Player
1,935
13th of 106
Single-core only

Rank among the 106 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 94.8 GB/s 3 1.46%
Scale higher is better 74.2 GB/s 3 0.594%
Add higher is better 75.3 GB/s 3 0.189%
Triad higher is better 75.5 GB/s 3 0.461%

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 649/s 3 0.411%
RSA 4096 verifies higher is better 43,052/s 3 0.0636%
SHA-256 higher is better 2.96 GB/s 3 0.929%
AES-256-GCM higher is better 8.95 GB/s 3 0.147%
ChaCha20-Poly1305 higher is better 5.38 GB/s 3 0.008%

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 7.18 ms 3 0.257%
Requests per second, 90th percentile higher is better 548/s 3 0.211%
Request latency, 99.9th percentile lower is better 12.2 ms 3 2%
Requests per second, average higher is better 543/s 3 0.217%

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

Disk

Block size Read Write Total
4K 256 MB/s (65,547 IOPS) 256.7 MB/s (65,720 IOPS) 512.8 MB/s (131,267 IOPS)
64K 2.27 GB/s (37,223 IOPS) 2.28 GB/s (37,419 IOPS) 4.56 GB/s (74,642 IOPS)
512K 3.69 GB/s (7,556 IOPS) 3.89 GB/s (7,958 IOPS) 7.58 GB/s (15,514 IOPS)
1M 3.96 GB/s (4,053 IOPS) 4.22 GB/s (4,323 IOPS) 8.18 GB/s (8,376 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 4.41 Gbps 4.24 Gbps
Eranium, Amsterdam, NL (100G) IPv4 5.66 Gbps 3.88 Gbps
Leaseweb, Singapore, SG (10G) IPv4 203 Mbps 1.29 Gbps
Clouvider, Los Angeles, CA, US (10G) IPv4 752 Mbps 1.19 Gbps
Leaseweb, NYC, NY, US (10G) IPv4 1.38 Gbps 2.26 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 8,870 0.113 ms
fdatasync 7,804 0.128 ms
fsync 6,145 0.163 ms
fsync_writethrough Not supported here n/a
open_sync 6,665 0.15 ms
Two 8 kB writes
open_datasync 4,230 0.236 ms
fdatasync 7,689 0.13 ms
fsync 6,439 0.155 ms
fsync_writethrough Not supported here n/a
open_sync 3,373 0.296 ms
open_sync at different write sizes
1 * 16kB open_sync write 6,517 0.153 ms
2 * 8kB open_sync writes 3,259 0.307 ms
4 * 4kB open_sync writes 1,666 0.6 ms
8 * 2kB open_sync writes 841 1.19 ms
16 * 1kB open_sync writes 435 2.3 ms
fsync on a non-write descriptor
write, fsync, close 5,986 0.167 ms
write, close, fsync 5,636 0.177 ms
Non-synced 8 kB writes
write 662,923 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 2,499 0.39 ms 0.453 ms 0.599 ms 2.41 ms 0
Read/write (TPC-B like) 8 9,203 0.763 ms 1.51 ms 2.19 ms 5.9 ms 0
Read/write (TPC-B like) 32 7,708 3.16 ms 10.2 ms 17.9 ms 30 ms 0
Read only 1 28,056 0.036 ms 0.044 ms 0.052 ms 0.081 ms 0
Read only 8 113,836 0.063 ms 0.114 ms 0.144 ms 0.196 ms 0
Read only 32 119,086 0.232 ms 0.349 ms 0.449 ms 0.646 ms 0

Scale 10 (1,000,000 accounts), loaded in 0.93 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 2,271 0.428 ms 0.509 ms 0.908 ms 2.48 ms 0
Read/write (TPC-B like) 8 9,335 0.793 ms 1.33 ms 2.25 ms 3.77 ms 0
Read/write (TPC-B like) 32 9,561 3.16 ms 5.27 ms 6.94 ms 11 ms 0
Read only 1 23,814 0.041 ms 0.054 ms 0.065 ms 0.096 ms 0
Read only 8 88,778 0.082 ms 0.152 ms 0.192 ms 0.4 ms 0
Read only 32 89,334 0.315 ms 0.511 ms 0.638 ms 1.16 ms 0

Scale 100 (10,000,000 accounts), loaded in 8.72 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 1 475 9,497 2 ms 2.62 ms 4.41 ms 0
Read/write 8 1,167 23,343 6.55 ms 11 ms 15 ms 0
Read/write 32 1,755 35,095 17.6 ms 29.7 ms 36.9 ms 0
Read only 1 1,187 18,986 0.826 ms 0.971 ms 1.25 ms 0
Read only 8 4,397 70,348 1.79 ms 2.18 ms 2.66 ms 0
Read only 32 3,986 63,779 7.99 ms 10.1 ms 11.7 ms 0

10 tables of 100,000 rows, loaded in 4.3 s. 60 s per run, one connection per thread.

Redis

memtier

Mix Operations/s Hits/s Median p99 p99.9
1:10 set:get 63,696 57,905 0.983 ms 1.93 ms 2.16 ms
1:1 set:get 63,448 31,724 0.991 ms 1.94 ms 2.13 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 1,162 4,233 6,359 372 ms
Express express-postgresjs 13,292 19,999 60,183 37.7 ms
Fastify fastify-postgres 12,965 20,675 60,082 38 ms
NestJS nestjs 10,257 11,444 56,453 42.4 ms
Nuxt nuxt 11,882 19,456 57,358 49.8 ms
FastAPI fastapi 7,672 9,740 34,720 221 ms
Django django-postgresql 3,504 3,972 10,504 311 ms
Rails rails 5,384 7,486 34,499 70.2 ms
Laravel laravel 1,368 1,430 3,003 203 ms
Symfony symfony 1,721 1,897 6,355 165 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.

WooCommerce

Page mix

Step Requests/s Median p95 p99 Error rate
8 virtual users
Home page 3.95 404 ms 536 ms 594 ms 0%
Shop page 3.95 532 ms 713 ms 829 ms 0%
Category page 3.95 361 ms 571 ms 696 ms 0%
Product page 3.95 376 ms 497 ms 536 ms 0%
Product search 3.95 338 ms 442 ms 479 ms 0%
32 virtual users
Home page 3.7 1.64 s 1.87 s 1.97 s 0%
Shop page 3.7 2.13 s 2.46 s 2.56 s 0%
Category page 3.7 1.66 s 2.19 s 2.36 s 0%
Product page 3.7 1.59 s 1.81 s 1.88 s 0%
Product search 3.7 1.46 s 1.66 s 1.78 s 0%

Cart and checkout

Step Requests/s Median p95 p99 Error rate
8 virtual users
Add to cart 7.63 418 ms 567 ms 610 ms 0%
Cart 7.63 347 ms 466 ms 490 ms 0%
Checkout 7.63 281 ms 374 ms 408 ms 0%
32 virtual users
Add to cart 7.13 1.65 s 1.95 s 2.05 s 0%
Cart 7.13 1.56 s 1.78 s 1.87 s 0%
Checkout 7.13 1.26 s 1.49 s 1.55 s 0%

k6 with a fixed number of virtual users, 61 to 66 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 2.72 163 ms 199 ms 413 ms 0%
Topic JSON 1.37 82.2 ms 96.5 ms 136 ms 0%
Home page 1.37 145 ms 179 ms 343 ms 0%
Categories page 0.676 167 ms 205 ms 273 ms 0%
User activity page 0.676 104 ms 119 ms 166 ms 0%
4 virtual users
Topic page 7.03 249 ms 331 ms 644 ms 0%
Topic JSON 3.51 122 ms 172 ms 240 ms 0%
Home page 3.51 225 ms 326 ms 478 ms 0%
Categories page 1.79 257 ms 393 ms 768 ms 0%
User activity page 1.75 159 ms 369 ms 508 ms 0%
16 virtual users
Topic page 5.96 1.11 s 1.53 s 1.85 s 0%
Topic JSON 2.98 810 ms 1.1 s 1.27 s 0%
Home page 3 1.06 s 1.4 s 2.11 s 0%
Categories page 1.51 1.1 s 1.71 s 2.02 s 0%
User activity page 1.47 861 ms 1.2 s 1.44 s 0%

Anonymous

Step Requests/s Median p95 p99 Error rate
16 virtual users
Topic page 134 29 ms 59.2 ms 80.8 ms 0%
Topic JSON 66.9 27.4 ms 57.9 ms 81.6 ms 0%
Home page 66.9 29.7 ms 60.1 ms 84.9 ms 0%
Categories page 33.5 28.8 ms 59.3 ms 81.4 ms 0%
User activity page 33.5 177 ms 254 ms 422 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 4096 MB of Postgres shared buffers, what its installer picks for a 16 GB box. Its installer adds no swap on a box this big. The container peaked at 2,750 MB.

Ghost

Readers

Step Requests/s Median p95 p99 Error rate
1 virtual user
Home page 11.6 23.4 ms 26.5 ms 29.2 ms 0%
Post page 28.9 14.7 ms 17.5 ms 20.9 ms 0%
Tag page 5.78 15.8 ms 18.3 ms 22.3 ms 0%
Author page 5.78 16.2 ms 19 ms 22.6 ms 0%
RSS feed 5.76 18.3 ms 22 ms 29.7 ms 0%
4 virtual users
Home page 14.1 73.2 ms 90.1 ms 102 ms 0%
Post page 35.2 55 ms 71.5 ms 78.4 ms 0%
Tag page 7.06 46.8 ms 61.1 ms 67.8 ms 0%
Author page 7.04 46.2 ms 59.1 ms 67.6 ms 0%
RSS feed 7.03 49.2 ms 65.1 ms 70.6 ms 0%
16 virtual users
Home page 14.7 275 ms 318 ms 357 ms 0%
Post page 36.7 231 ms 272 ms 309 ms 0%
Tag page 7.33 151 ms 181 ms 209 ms 0%
Author page 7.38 150 ms 184 ms 219 ms 0%
RSS feed 7.3 154 ms 185 ms 203 ms 0%

Staff in Ghost Admin

Step Requests/s Median p95 p99 Error rate
1 virtual user
Post list 6.15 23.4 ms 27.3 ms 33.8 ms 0%
Post editor 18.5 10.4 ms 13.8 ms 16 ms 0%
Saving a draft 18.5 24.9 ms 29.6 ms 35.4 ms 0%
Member list 6.15 26 ms 30.7 ms 36.2 ms 0%
4 virtual users
Post list 7.89 33.1 ms 37.1 ms 44.1 ms 0%
Post editor 23.7 18.7 ms 22.6 ms 25.2 ms 0%
Saving a draft 23.8 122 ms 146 ms 161 ms 0%
Member list 7.93 32.3 ms 36.2 ms 38.4 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 552 MB and MySQL at 625 MB.

Wagtail

Visitors

Step Requests/s Median p95 p99 Error rate
1 virtual user
Home page 2.14 62.6 ms 71.6 ms 75.6 ms 0%
Blog index 2.14 198 ms 222 ms 227 ms 0%
Blog post 6.39 21.6 ms 31.9 ms 36.7 ms 0%
Bread page 6.41 23.4 ms 27.2 ms 29 ms 0%
Location page 2.14 21.6 ms 25.9 ms 28 ms 0%
Site search 2.12 35 ms 59.7 ms 65.1 ms 0%
4 virtual users
Home page 2.84 209 ms 290 ms 311 ms 0%
Blog index 2.85 708 ms 1.02 s 1.14 s 0%
Blog post 8.5 56.6 ms 90.2 ms 133 ms 0%
Bread page 8.53 59.4 ms 91 ms 99.4 ms 0%
Location page 2.85 57.5 ms 88.4 ms 131 ms 0%
Site search 2.84 81.7 ms 197 ms 229 ms 0%
16 virtual users
Home page 3.17 583 ms 807 ms 945 ms 0%
Blog index 3.26 1.54 s 1.8 s 1.91 s 0%
Blog post 9.51 332 ms 525 ms 681 ms 0%
Bread page 9.55 323 ms 542 ms 662 ms 0%
Location page 3.2 308 ms 481 ms 535 ms 0%
Site search 3.18 403 ms 614 ms 830 ms 0%

Editors in the Wagtail admin

Step Requests/s Median p95 p99 Error rate
1 virtual user
Admin dashboard 0.86 71.5 ms 84.7 ms 93.2 ms 0%
Page explorer 1.75 265 ms 303 ms 445 ms 0%
Page editor 2.6 131 ms 152 ms 210 ms 0%
Image library 0.86 51.5 ms 56.5 ms 400 ms 0%
Admin page search 0.86 30.2 ms 62.7 ms 186 ms 0%
4 virtual users
Admin dashboard 1.22 285 ms 500 ms 577 ms 0%
Page explorer 2.48 636 ms 1.14 s 1.31 s 0%
Page editor 3.74 431 ms 754 ms 835 ms 0%
Image library 1.24 214 ms 463 ms 531 ms 0%
Admin page search 1.22 97.1 ms 293 ms 382 ms 0%

k6 with a fixed number of virtual users, 60 to 61 s per level, after a 20 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 298 MB, Postgres at 148 MB, and Redis at 8 MB.

Advertised vs. measured

Advertised Measured
vCPU 4 4
CPU Dedicated vCPU 0% of busy CPU time stolen
RAM 16 GB 15.2 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-Milan Processor
CPU cores
4
CPU frequency
2197 MHz
AES-NI
Yes
VM-x/AMD-V
No
RAM reported
15.2 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
CCX23
Product line
general_purpose
Location
Helsinki DC Park 1
vCPU
4 dedicated vCPU
Architecture
x86_64
RAM
16 GB
Storage
160 GB NVMe
Virtualization
KVM
Network
Unknown
Transfer
21.99 TB
Port 25
On request
Monthly price at test time
$101.49
Hourly price at test time
$0.16
YABS version
v2026-09-20
Ran at
Oct 12, 2026, 01:08:37 UTC
Parser version
5

Run timeline

Queued
Oct 12, 2026, 01:06:45 UTC
Server created
Oct 12, 2026, 01:08:31 UTC
First update from the server
Oct 12, 2026, 01:08:31 UTC
Completed
Oct 12, 2026, 02:40:29 UTC
Server deleted
Oct 12, 2026, 02:40:29 UTC
Cost of this run
$0.25
Watch party 0 watching

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

Watch party

Loading the party