• There is NO official Otland's Discord server and NO official Otland's server list. The Otland's Staff does not manage any Discord server or server list. Moderators or administrator of any Discord server or server lists have NO connection to the Otland's Staff. Do not get scammed!
  • 2026 staff recruitment is open! Check it out and consider applying!

VPS/dedic benchmarks

Added 3 new benchmarks of DigitalOcean.com cloud servers (not recommended):
docs.google.com

VPS Benchmarks

docs.google.com docs.google.com

Cloud (VPS) servers cost more than some dedics in OVH/Hetzner and they are much slower. Also results are very unstable, sometimes I get 50% lower score in CPU/RAM test than the best test score on given machine. Also on cheapest option (48$ for 4 core and 8 GB RAM!!!) with SSD response time sometimes jump to 10 ms.

EDIT:
I tried to order server with 'dedicated CPU' (not shared), but it's not available without sending 'increase resources request'.
I've send one with information that I want to benchmark machine with 4 cores and 8 GB RAM, because I'm looking for game server hosting.
They DENIED my request!
While we've gone ahead and carefully reviewed your account, unfortunately, your account does not qualify for the requested limit based on our internal assessment guidelines.
Never heard about company that does not allow users to buy so slow servers.
I've asked for 109$ server (4 core, 8 GB RAM) with specification that I could buy in Hetzner.com for 30$ (4 dedicated core, 16 GB RAM).
 
Last edited:
HP Thin Client T640(ryzen)

AMD Ryzen Embedded R1505G with Radeon Vega Gfx (810F81)
8GB RAM
256 SSD NVME

CPU Freq: - - - - - - - - -

1.1 CPU - 7zip

3336 1 core

8667 4 core

1.2 CPU - sysbench

sysbench 1 core 1625

sysbench 4 core 3445

2.1 RAM - transfer

1024 1 core - 4 core

2.2 RAM - operations per second

1048576 1 core - 4 core

4.1 dd - write transfer

4294967296 bytes (4,3 GB, 4,0 GiB) copied, 5,95765 s, 721 MB/s

4.2 dd - read transfer

4294967296 bytes (4,3 GB, 4,0 GiB) copied, 2,51598 s, 1,7 GB/s

4.3 fio - write IOPS [operations per second]

write iops 1043190

4.4 fio - read IOPS [operations per second]

read iops 1018550

4.5 ioping - latency

min/avg/max/mdev = 127.1 us / 7.38 ms / 7.65 ms / 958.0 us
 
Last edited:
TibiCAM said:
Have you tried Hetzner's ARM64 cloud servers? It seems their CAX21 might be a good option.

4 vCPU -- 8 GB RAM -- 80 GB Disk Space -- 20 TB bandwidth

And depending on country it's around 8€/month.

I read a good benchmark / article about it here: The performance review of Hetzner's CAX-line ARM64 servers and the practical experience of WebP Cloud Services on them. (https://blog.webp.se/hetzner-arm64-en/)

Might get myself one of those.
Click to expand...
I've tested 16 vCore ARM CPU for 29 EUR. Results are in 28 line of:
docs.google.com

VPS Benchmarks

docs.google.com docs.google.com
In lines 25-27 are Hetzner cloud servers with AMD vCore (shared) and dedicated core (more expensive) CPUs for comparison. They cost around 2 times more and they are only a bit faster.

There are now available new Intel CPU cloud servers in Hetzner. 4 cores with 8 GB RAM costs only 7.75 EUR.
 
I use hetzner services using 8gb ram and 4core cpu you get basic quality
if you go with 2vcpu (shared) and 4gb ram that thing works slower than a pentium 4 and cant even handle tfs without lagging :p so the 8gb option is a minimum that I would recommend unless you trying to host some node apps or small web app
 
it would be good if you could tell us a little about the servers in the west that are hosted in canada, usa or brazil.
 
Odisk said:
it would be good if you could tell us a little about the servers in the west that are hosted in canada, usa or brazil.
Click to expand...
OVH has datacenters in Canada and USA. There are the same servers as in Europe. To order OVH server in USA you may need to use US site and create separate OVH account on US site (probably tax related): OVHcloud (https://us.ovhcloud.com/)

Hetzner has datacenters in USA (east/west), but there are only available Cloud servers (no dedics) and only with AMD CPUs. They are cheaper and faster than OVH, but Hetzner anti-ddos is very weak.

IDK anyone who host server in Brazil. There are many brazilian companies that just resell OVH US servers.
I've checked 7 top BR servers from otservlist:

Some OTS servers reported as OVH US east are hosted in Hetzner US and use OVH VPS servers as proxy servers to block DDoS attacks, so their IP on otservlist points to OVH, but they are hosted on cheaper and faster servers in Hetzner.
 
zorrasss said:
A good thing about hetzner is that the price is monthly, but they will only bill per hour rate, so its great for testing purposes
Click to expand...
I use their servers all time for tests.

Need to compile some random OTS engine on Linux and fix compilation issues? 16 cores, 32 GB RAM with 2 gb/s network for 0.1$ per hour (0.07$ if Intel CPU servers are available). Compilation on that machine takes 10-30 seconds.
Hetzner delivers Cloud servers in 15-30 seconds. It's much faster than setting up docker to work with given OS and finding out what C++ libraries/packages to install.
There are many other companies that deliver similar servers (AWS, Google Cloud, Microsoft Azure), but their cloud servers often come with a lot of configuration required, to make them run like VPSes. Hetzner Cloud servers are configured like VPSes (full access to everything by default).
 
Gesior.pl said:
I use their servers all time for tests.

Need to compile some random OTS engine on Linux and fix compilation issues? 16 cores, 32 GB RAM with 2 gb/s network for 0.1$ per hour (0.07$ if Intel CPU servers are available). Compilation on that machine takes 10-30 seconds.
Hetzner delivers Cloud servers in 15-30 seconds. It's much faster than setting up docker to work with given OS and finding out what C++ libraries/packages to install.
There are many other companies that deliver similar servers (AWS, Google Cloud, Microsoft Azure), but their cloud servers often come with a lot of configuration required, to make them run like VPSes. Hetzner Cloud servers are configured like VPSes (full access to everything by default).
Click to expand...
If it comes to host a productive server on south america, do you have any thoughts on it? Anti-DDOS wise.
 
Gesior.pl said:
OVH has datacenters in Canada and USA. There are the same servers as in Europe. To order OVH server in USA you may need to use US site and create separate OVH account on US site (probably tax related): OVHcloud (https://us.ovhcloud.com/)

Hetzner has datacenters in USA (east/west), but there are only available Cloud servers (no dedics) and only with AMD CPUs. They are cheaper and faster than OVH, but Hetzner anti-ddos is very weak.

IDK anyone who host server in Brazil. There are many brazilian companies that just resell OVH US servers.
I've checked 7 top BR servers from otservlist:
Click to expand...
for ovh you don't need anything... when you go to rent a server it asks you where you want it to be europe usa canada etc and the south american servers are a bit expensive, but neep is one of the best servers example rubinot is there and it is a “stable” server.
 
Odisk said:
for ovh you don't need anything... when you go to rent a server it asks you where you want it to be europe usa canada etc
Click to expand...
You can order server in Europe from USA account, but you can't order server in USA from account made on European OVH website - you can order VPS in Canada, but not USA. It's US tax related problem.
Odisk said:
neep is one of the best servers example rubinot is there and it is a “stable” server.
Click to expand...
It's stable, but you have to write pretty complicated software on your server to use Neep anti-DDoS effectively. Your VPS/dedic has to list all allowed IPs - all players IPs - before they can connect to VPS/dedic at all. RubinOT probably uses some website that Tibia Client opens to list players IPs (ex. login.php for 12+ client).


VPS CPU results listed in this thread are just tests ran once/few times. Cheap VPSes sometimes get much slower for few minutes.. in my tests, up to 13 TIMES slower!
I did some 50+ days tests of CPU speed using OVH, Hetzner and Contabo servers.

Results and summary of CPU usage:

HETZNER CX33 (4 cores, 8 GB RAM, 7.7 EUR):
There were some single core CPU speed drops making it up to 50% slower, but they were rare.
Code: 
sysbench-single: 38711 entries, Min: 464.00, Max: 911.37, Avg: 831.45
sysbench-multi: 38711 entries, Min: 2451.77, Max: 3569.41, Avg: 3252.31
7zip-single: 38711 entries, Min: 1878.00, Max: 2777.00, Avg: 2469.26
7zip-multi: 38711 entries, Min: 7180.00, Max: 11441.00, Avg: 9565.10

CONTABO Cloud VPS 10 (3 cores, 8 GB RAM, 5.5 EUR):
Min: 109.99 sometimes CPU gets 14 times slower than average Avg: 1469.52!
Maybe it was slower or maybe it was frozen for some time ex. it worked full speed for 0.1 sec, then do nothing for 0.9 sec - horrible CPU for OTS.
Code: 
sysbench-single: 27612 entries, Min: 109.99, Max: 1564.79, Avg: 1469.52
sysbench-multi: 27612 entries, Min: 2474.94, Max: 4687.85, Avg: 4193.35
7zip-single: 27611 entries, Min: 1608.00, Max: 3673.00, Avg: 2361.31
7zip-multi: 27609 entries, Min: 4457.00, Max: 12822.00, Avg: 8039.30

OVH VPS-1 (4 cores, 8 GB RAM, 5.5 EUR):
Pretty stable results, but.. I ordered VPS-1 and get CPU with 2.99 GHz clock, my friend ordered VPS-5 and get CPU with 2.4 GHz clock.
By clock speed, it should be "only" 25% faster, but in all single core tests it was 35-50% faster!
You can pay the same price and get randomly 30-50% slower/faster CPU! If you plan to order VPS in OVH, do not order it for more than 1 month. You can cancel it after a month and order another hoping to get faster machine. Check your CPU clock speed using cat /proc/cpuinfo | grep 'MHz'.
Code: 
sysbench-single: 54784 entries, Min: 912.13, Max: 1126.84, Avg: 1057.65
sysbench-multi: 54783 entries, Min: 3253.02, Max: 4340.53, Avg: 4268.30
7zip-single: 54781 entries, Min: 3201.00, Max: 4107.00, Avg: 3685.39
7zip-multi: 54781 entries, Min: 12277.00, Max: 16610.00, Avg: 15749.31

I added these server tests to my spreadsheet - tests made minutes after delivery:
docs.google.com

VPS Benchmarks

docs.google.com docs.google.com
 
hello @Gesior.pl What would benchmarks look like on a Raspberry Pi (3/4/5)? Which server from the list would it be similar to? Maybe someone has one at home and could run a test script?
 
check said:
Maybe someone has one at home and could run a test script?
Click to expand...
I have Raspberry Pi 3 B v1.2. I will try to run benchmarks on weekend.
Some tests say that Pi 5 CPU is around 4 times faster than Pi 3 (single and multi core). Anyone has Pi 5 at home to test?

There are multiple tests results available online, but they use custom parameters to sysbench and they don't show results from terminal, just some number like 10k for single core (15% more than fastest CPU I've tested - AMD Ryzen 7950X3D CPU -, impossible).
I don't expect results comparable with any VPS. OVH X-XX years ago tried to host 'cheap dedics' (kimsufi.com, now part of ovh.com site) on ARM based machines (like Raspberry Pi) for ~3 EUR per month - mostly with big HDDs to store backups on them, but they resigned after first 'edition'.
 
Gesior.pl said:
I have Raspberry Pi 3 B v1.2. I will try to run benchmarks on weekend.
Some tests say that Pi 5 CPU is around 4 times faster than Pi 3 (single and multi core). Anyone has Pi 5 at home to test?

There are multiple tests results available online, but they use custom parameters to sysbench and they don't show results from terminal, just some number like 10k for single core (15% more than fastest CPU I've tested - AMD Ryzen 7950X3D CPU -, impossible).
I don't expect results comparable with any VPS. OVH X-XX years ago tried to host 'cheap dedics' (kimsufi.com, now part of ovh.com site) on ARM based machines (like Raspberry Pi) for ~3 EUR per month - mostly with big HDDs to store backups on them, but they resigned after first 'edition'.
Click to expand...

Not benchmarking disk because it's useless.

Kernel:
6.8.0-1040-raspi #44-Ubuntu SMP PREEMPT_DYNAMIC Wed Sep 24 18:30:35 UTC 2025 aarch64 aarch64 aarch64 GNU/Linux.
OS: Ubuntu 24.04.3 LTS

sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)

Raspberry Pi 4 Model B Rev 1.5 (4GB RAM)

CPU:
Code: 
Number of threads: 1
Prime numbers limit: 10000

CPU speed:
    events per second:  1748.09

General statistics:
    total time:                          10.0004s
    total number of events:              17489

Latency (ms):
         min:                                    0.57
         avg:                                    0.57
         max:                                    0.88
         95th percentile:                        0.60
         sum:                                 9993.93

Threads fairness:
    events (avg/stddev):           17489.0000/0.00
    execution time (avg/stddev):   9.9939/0.00

Number of threads: 2
Prime numbers limit: 10000

CPU speed:
    events per second:  3495.72

General statistics:
    total time:                          10.0006s
    total number of events:              34974

Latency (ms):
         min:                                    0.57
         avg:                                    0.57
         max:                                    0.96
         95th percentile:                        0.58
         sum:                                19989.97

Threads fairness:
    events (avg/stddev):           17487.0000/13.00
    execution time (avg/stddev):   9.9950/0.00


Number of threads: 4
Prime numbers limit: 10000

CPU speed:
    events per second:  6985.02

General statistics:
    total time:                          10.0006s
    total number of events:              69884

Latency (ms):
         min:                                    0.57
         avg:                                    0.57
         max:                                    1.15
         95th percentile:                        0.58
         sum:                                39982.03

Threads fairness:
    events (avg/stddev):           17471.0000/44.38
    execution time (avg/stddev):   9.9955/0.00

Number of threads: 16
Prime numbers limit: 10000

CPU speed:
    events per second:  6958.35

General statistics:
    total time:                          10.0026s
    total number of events:              69621

Latency (ms):
         min:                                    0.57
         avg:                                    2.29
         max:                                   23.59
         95th percentile:                        9.56
         sum:                               159637.59

Threads fairness:
    events (avg/stddev):           4351.3125/548.25
    execution time (avg/stddev):   9.9773/0.02

RAM transfer
Code: 
Number of threads: 1

Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 ( 3753.13 per second)

1024.00 MiB transferred (3753.13 MiB/sec)

General statistics:
    total time:                          0.2684s
    total number of events:              1024

Latency (ms):
         min:                                    0.22
         avg:                                    0.26
         max:                                    0.43
         95th percentile:                        0.30
         sum:                                  266.25

Threads fairness:
    events (avg/stddev):           1024.0000/0.00
    execution time (avg/stddev):   0.2662/0.00

Number of threads: 2

Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 ( 4365.34 per second)

1024.00 MiB transferred (4365.34 MiB/sec)

General statistics:
    total time:                          0.2303s
    total number of events:              1024

Latency (ms):
         min:                                    0.19
         avg:                                    0.45
         max:                                    0.70
         95th percentile:                        0.58
         sum:                                  457.01

Threads fairness:
    events (avg/stddev):           512.0000/0.00
    execution time (avg/stddev):   0.2285/0.00

Number of threads: 4

Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 ( 4995.54 per second)

1024.00 MiB transferred (4995.54 MiB/sec)

General statistics:
    total time:                          0.2006s
    total number of events:              1024

Latency (ms):
         min:                                    0.18
         avg:                                    0.78
         max:                                    1.61
         95th percentile:                        1.06
         sum:                                  794.04

Threads fairness:
    events (avg/stddev):           256.0000/0.00
    execution time (avg/stddev):   0.1985/0.00

Number of threads: 16

Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 ( 4896.44 per second)

1024.00 MiB transferred (4896.44 MiB/sec)

General statistics:
    total time:                          0.2060s
    total number of events:              1024

Latency (ms):
         min:                                    0.56
         avg:                                    2.76
         max:                                   16.19
         95th percentile:                       12.75
         sum:                                 2823.44

Threads fairness:
    events (avg/stddev):           64.0000/0.00
    execution time (avg/stddev):   0.1765/0.03

RAM OPS
Code: 
Number of threads: 1

Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (2343384.99 per second)

1024.00 MiB transferred (2288.46 MiB/sec)

General statistics:
    total time:                          0.4432s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    0.07
         95th percentile:                        0.00
         sum:                                  209.46

Threads fairness:
    events (avg/stddev):           1048576.0000/0.00
    execution time (avg/stddev):   0.2095/0.00

Number of threads: 2

Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (4444002.83 per second)

1024.00 MiB transferred (4339.85 MiB/sec)

General statistics:
    total time:                          0.2318s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    0.08
         95th percentile:                        0.00
         sum:                                  230.68

Threads fairness:
    events (avg/stddev):           524288.0000/0.00
    execution time (avg/stddev):   0.1153/0.00

Number of threads: 4

Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (8425646.60 per second)

1024.00 MiB transferred (8228.17 MiB/sec)

General statistics:
    total time:                          0.1203s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    0.36
         95th percentile:                        0.00
         sum:                                  239.92

Threads fairness:
    events (avg/stddev):           262144.0000/0.00
    execution time (avg/stddev):   0.0600/0.00

Number of threads: 16

Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (7944663.21 per second)

1024.00 MiB transferred (7758.46 MiB/sec)


General statistics:
    total time:                          0.1277s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                   17.02
         95th percentile:                        0.00
         sum:                                  474.12

Threads fairness:
    events (avg/stddev):           65536.0000/0.00
    execution time (avg/stddev):   0.0296/0.01

Raspberry Pi 5 Model B Rev 1.0 (8GB RAM)

CPU:
Code: 
Number of threads: 1
Prime numbers limit: 10000

CPU speed:
    events per second:  2720.83

General statistics:
    total time:                          10.0001s
    total number of events:              27213

Latency (ms):
         min:                                    0.37
         avg:                                    0.37
         max:                                    0.49
         95th percentile:                        0.37
         sum:                                 9995.46

Threads fairness:
    events (avg/stddev):           27213.0000/0.00
    execution time (avg/stddev):   9.9955/0.00

Number of threads: 2
Prime numbers limit: 10000

CPU speed:
    events per second:  5442.67

General statistics:
    total time:                          10.0004s
    total number of events:              54438

Latency (ms):
         min:                                    0.37
         avg:                                    0.37
         max:                                    0.67
         95th percentile:                        0.37
         sum:                                19990.89

Threads fairness:
    events (avg/stddev):           27219.0000/2.00
    execution time (avg/stddev):   9.9954/0.00

Number of threads: 4
Prime numbers limit: 10000

CPU speed:
    events per second: 10877.04

General statistics:
    total time:                          10.0003s
    total number of events:              108792

Latency (ms):
         min:                                    0.37
         avg:                                    0.37
         max:                                    8.42
         95th percentile:                        0.37
         sum:                                39980.21

Threads fairness:
    events (avg/stddev):           27198.0000/50.29
    execution time (avg/stddev):   9.9951/0.00

Number of threads: 16
Prime numbers limit: 10000

CPU speed:
    events per second: 10877.76

General statistics:
    total time:                          10.0013s
    total number of events:              108804

Latency (ms):
         min:                                    0.37
         avg:                                    1.47
         max:                                   44.38
         95th percentile:                        9.39
         sum:                               159610.48

Threads fairness:
    events (avg/stddev):           6800.2500/111.23
    execution time (avg/stddev):   9.9757/0.02
RAM transfer:
Code: 
Number of threads: 1
Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 (11428.11 per second)

1024.00 MiB transferred (11428.11 MiB/sec)


General statistics:
    total time:                          0.0879s
    total number of events:              1024

Latency (ms):
         min:                                    0.08
         avg:                                    0.08
         max:                                    0.17
         95th percentile:                        0.10
         sum:                                   87.04

Threads fairness:
    events (avg/stddev):           1024.0000/0.00
    execution time (avg/stddev):   0.0870/0.00

Number of threads: 2
Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 (21928.39 per second)

1024.00 MiB transferred (21928.39 MiB/sec)


General statistics:
    total time:                          0.0450s
    total number of events:              1024

Latency (ms):
         min:                                    0.08
         avg:                                    0.09
         max:                                    0.19
         95th percentile:                        0.10
         sum:                                   88.87

Threads fairness:
    events (avg/stddev):           512.0000/0.00
    execution time (avg/stddev):   0.0444/0.00

Number of threads: 4
Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 (36670.40 per second)

1024.00 MiB transferred (36670.40 MiB/sec)


General statistics:
    total time:                          0.0262s
    total number of events:              1024

Latency (ms):
         min:                                    0.08
         avg:                                    0.10
         max:                                    0.47
         95th percentile:                        0.20
         sum:                                  103.10

Threads fairness:
    events (avg/stddev):           256.0000/0.00
    execution time (avg/stddev):   0.0258/0.00

Number of threads: 16
Running memory speed test with the following options:
  block size: 1024KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1024 (34007.55 per second)

1024.00 MiB transferred (34007.55 MiB/sec)


General statistics:
    total time:                          0.0284s
    total number of events:              1024

Latency (ms):
         min:                                    0.08
         avg:                                    0.12
         max:                                    6.25
         95th percentile:                        0.25
         sum:                                  126.67

Threads fairness:
    events (avg/stddev):           64.0000/0.00
    execution time (avg/stddev):   0.0079/0.00

RAM OPS:
Code: 
Number of threads: 1
Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (3597542.12 per second)

1024.00 MiB transferred (3513.22 MiB/sec)


General statistics:
    total time:                          0.2898s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    0.01
         95th percentile:                        0.00
         sum:                                  145.49

Threads fairness:
    events (avg/stddev):           1048576.0000/0.00
    execution time (avg/stddev):   0.1455/0.00

Number of threads: 2
Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (4975656.38 per second)

1024.00 MiB transferred (4859.04 MiB/sec)


General statistics:
    total time:                          0.2091s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    0.01
         95th percentile:                        0.00
         sum:                                  256.50

Threads fairness:
    events (avg/stddev):           524288.0000/0.00
    execution time (avg/stddev):   0.1283/0.00

Number of threads: 4
Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (8991734.95 per second)

1024.00 MiB transferred (8780.99 MiB/sec)


General statistics:
    total time:                          0.1149s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    0.04
         95th percentile:                        0.00
         sum:                                  285.93

Threads fairness:
    events (avg/stddev):           262144.0000/0.00
    execution time (avg/stddev):   0.0715/0.00

Number of threads: 16
Running memory speed test with the following options:
  block size: 1KiB
  total size: 1024MiB
  operation: write
  scope: global

Total operations: 1048576 (8897476.38 per second)

1024.00 MiB transferred (8688.94 MiB/sec)


General statistics:
    total time:                          0.1163s
    total number of events:              1048576

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                   18.00
         95th percentile:                        0.00
         sum:                                  851.72

Threads fairness:
    events (avg/stddev):           65536.0000/0.00
    execution time (avg/stddev):   0.0532/0.02
 
godoyxd said:
Not benchmarking disk because it's useless.
Click to expand...
@check
Added Raspberry Pi 4 and 5 to VPS results:
docs.google.com

VPS Benchmarks

docs.google.com docs.google.com
You can compare Pi lines 39 and 40 with cheap VPSes in lines 34-36.
Raspberry Pi 5 is comparable with slow OVH.com VPS (line 37), but you can get (randomly, check my previous post) faster VPS in OVH, which is 50% faster than Pi 5. It's also around 2 times slower on single core than Intel 7700K from 2017, which you can rent for ~40 EUR/month.
 
I found a fantastic post on Reddit that compares pretty much all of Hetzner's VPS. Seems like CAX11 is the best budget still, even though they added a Intel/AMD server in October 2025 that is slightly cheaper.

https://www.reddit.com/r/hetzner/comments/1pano24/hetzner_cloud_server_benchmark_cx_vs_cax_vs_cpx

He included a ton of different programs, and the commands he used. Top post really. I don't know how to add a copy of it in here, so I will just link it. But maybe someone has time to put it here as well.
 
Back
Top