37
votes
Benchmarking of USB flash drive
Use built-in program disks from ubuntu launcher, select your pendrive, click on gear icon, and select benchmark option to run read/write speed test
- 505
20
votes
Running a simulation on pure Ubuntu vs on Ubuntu in Windows (WSL)
Your simulation software is most likely either CPU bound or memory bound. For such workloads, one would not except to see any significant difference between running the code on "bare metal" or inside ...
- 511
15
votes
Accepted
Are there concrete figures on the speed of bash vs dash?
This test is not representative for the boot process, but you can simply try out yourself by making a little test script, I called it shspeed:
$ cat shspeed
for a in `seq 10000`; do ( :; ); done
...
- 6,122
12
votes
Running a simulation on pure Ubuntu vs on Ubuntu in Windows (WSL)
Ubuntu in Windows (WSL - 2017 Fall Creators Update) is definitely slower than "Pure" Ubuntu in Linux environment.
For example screen painting takes many times longer in Windows 10 versus ...
- 102k
11
votes
Accepted
How can I run glmark2 on the dedicated GPU?
Launch the benchmark using the dGPU with the DRI_PRIME=1 option:
DRI_PRIME=1 glmark2
Edit : This answer is from mid 2018 when Xorg was the default session.
This means the solution is valid if the ...
- 31.2k
11
votes
Accepted
Disabling AMD's equivalent (on a Zen-1 Epyc) of Intel's "turbo boost" at runtime?
By default, AMD processors use the acpi-cpufreq CPU frequency scaling driver.
Check via:
grep . /sys/devices/system/cpu/cpufreq/policy*/scaling_driver
For that driver, to disable turbo boost, do:
...
- 15.4k
10
votes
Benchmarking of USB flash drive
I know it's a bit old of a post but! You could try the following:
sudo hdparm -Tt /dev/sd
I've used this on multiple occasions with good success. Nice and easy. Not to mention built in.
You could ...
- 346
8
votes
Is there a command-line equivalent to gnome-disks?
I cant speak to the old udisks benchmarking report but perhaps fio will be of use to you. fio is currently available for all versions of Ubuntu from Precise To Zesty
You can install it with sudo apt-...
- 36k
7
votes
Accepted
Benchmarking a HDD using the Benchmark utility in "Disks" on Ubuntu 16.04
hdparm is a standard CLI for testing disk speed and the drives don't have to be mounted (when doing read tests).
$ sudo hdparm -tT /dev/nvme0n1
/dev/nvme0n1:
Timing cached reads: 20760 MB in 1....
- 102k
7
votes
Running a simulation on pure Ubuntu vs on Ubuntu in Windows (WSL)
Think about it - in WSL your computer is running the full graphical Windows system (which is a horrific resource hog in the first place) plus the Ubuntu subsystem. In native Ubuntu it's only running ...
- 452
6
votes
How to interpret time "real", "user" and "sys"
Long story short, it simply means that your program didn't request to perform any privileged tasks, hence CPU has spend no time in the privileged (kernel) mode.
Basics First
First of all, there's ...
- 105k
5
votes
Accepted
Ubuntu Server 16.04: benchmarking disks from terminal
bonnie++ will do benchmarks, with many options for how to perform them.
It can be installed with sudo apt install bonnie++
- 22.7k
5
votes
Accepted
Does docker consume CPU the way VMs do?
Containers are no VMs. Nothing is virtualized, it all just runs as isolated processes on the host. Even the kernel is shared.
So the kind of CPU utilization you can observe from a container process ...
- 108k
4
votes
Measuring performance with shell
If you are using the bash or zsh shell, and only need a resolution of seconds, then you can use their SECONDS shell variable. From man bash:
SECONDS
Each time this parameter is referenced, ...
- 136k
4
votes
Measuring performance with shell
In addition to the time builtin, there exists /usr/bin/time, which is often more useful.
walt@bat:~(0)$ /usr/bin/time sleep 2
0.00user 0.00system 0:02.04elapsed 0%CPU (0avgtext+0avgdata ...
- 36.5k
4
votes
Accepted
Measuring performance with shell
Bash has a built in function for this called time. Just prepend it to any command and it will time how long the command takes to run. For more info, see help time :)
[user@sol ~]$ time sleep 2
real ...
- 2,380
4
votes
Can't run write benchmark in Disks (gnome-disks) on Ubuntu MATE 16.04.1
Note: Tested on 14.04 and 16.04
This is normal. You can't unmount a filesystem that is in use. On top of that the gnome-disks write benchmark is destructive and even warns you to backup before using ...
- 36k
4
votes
How to perform a detailed and quick 3D performance test
gfxbench
by Kishonti Ltd. is considered a golden standard in the industry.
You can download a free version of their benchmark at: https://gfxbench.com/linux-download/
Then run their scripts:
sh ...
4
votes
How can I run glmark2 on the dedicated GPU?
I'm on Xubuntu (22.04 Jammy).
My laptop has an AMD Integrated graphics chip, and a nVidia RTX3070 discrete graphics accelerator.
I have the proprietary nVidia drivers installed.
This worked for me:
...
- 151
3
votes
Accepted
Equivalent of hdparm -t for home directory?
You could use findmnt -T ex.
$ findmnt -T /home
TARGET SOURCE FSTYPE OPTIONS
/home /dev/sda6 ext4 rw,relatime,data=ordered
or to extract just the block device
$ findmnt -no SOURCE -T /home
/dev/...
- 136k
3
votes
Accepted
Can you run Compute Benchmark (GPU) using Geekbench 5 on 20.04?
It looks like you're using a demo version of Geekbench5. GPU benchmarking is a Geekbench5 Pro feature. Using the Pro version:
Run geekbench5 --compute-list to get a list of available GPU tests.
Run ...
- 78
2
votes
Accepted
Ubuntu's poor performance compared do Mac OS X (Same hardware!)
As you have confirmed in the comments, you had the use of Intel's "Procesor microcode firmware" driver disabled.
This microcode can be described as hot-patch for the firmware of Intel CPUs, which ...
- 108k
2
votes
Accepted
Looking for a way to halt background tasks/create a benchmark mode
One of the ways is to switch the system to multi-user target (this means without graphical session) reducing running processes to about 50:
sudo systemctl isolate multi-user.target
or even switch to ...
- 100k
1
vote
Accepted
Benchmark Ubuntu Persistent USB vs Installed System
Phoronix Test Suite (PTS) is a free and open-source benchmark software for Linux and other operating systems. The Phoronix Test Suite has been endorsed by sites such as Linux.com, LinuxPlanet and has ...
- 19.6k
1
vote
Measuring performance with shell
Fixing up your code
(assuming you're using Bash)
begin_time=$(date +%s) # Get seconds since Unix epoch.
sleep 2 # For example
echo $(($(date +%s) - begin_time)) seconds
This should output 2 ...
- 14.2k
1
vote
Accepted
sysbench anomaly: ubuntu server vs debian
The version of sysbench being used is different in each case. While the debain one is taking 1.3377 times longer per operation, it is also doing 2.6596 times the number of operations, for an expected ...
- 15.4k
1
vote
Accepted
Does any open source I/O benchmarking tool support ramping up IOPS?
Here's a bash script I cooked up (fio-ramp hosted at github):
#!/bin/bash
# Copyright 2019 Mikko Rantalainen
# License: MIT X License
#
# Debian/Ubuntu requirements:
# sudo apt install fio jq
#
# See ...
- 3,360
1
vote
Does any open source I/O benchmarking tool support ramping up IOPS?
fio has an option that discovers the highest IOPS that can be done under a certain latency... From the "I/O latency" section of the fio documentation:
latency_target=time
If set, fio will ...
- 312
1
vote
Accepted
Where can I find missed hpccinf.txt for hpcc?
According to man hpcc
The High Performance Computing (HPC) Challenge benchmark runs a suite of tests that
measure the performance of CPU, memory and network for HPC clusters. ...
- 100k
1
vote
Running a simulation on pure Ubuntu vs on Ubuntu in Windows (WSL)
I don't know whether this will affect your simulation in particular, but it might:
WSL does NOT use RAM for shared memory! It uses the disk!
This means, if your simulation uses shared memory (think /...
- 4,167
Only top scored, non community-wiki answers of a minimum length are eligible