Results of latency test, list of computers tested for use with LinuxCNC

More
01 Oct 2020 22:15 #184516 by PCW
Intel chips are normally fine for Mesa Ethernet cards,
but you do need to disable IRQ coalescing as described
in the hm2_eth man page.
The following user(s) said Thank You: seuchato

Please Log in or Create an account to join the conversation.

More
01 Oct 2020 22:46 #184520 by BeagleBrainz
Just a thought.
Whilst all this testing is great, the average forum user isn’t going to wade through this many pages. It would really need collating and summarising.
Maybe a new topic just listing the machine and latency with a link to the actual test in this thread maybe the go. Keep the summary on the first post, just add an edit for a new test.
I don’t think the wiki gets used that much, sometimes I think the docs get just as much use.
The following user(s) said Thank You: seuchato, txtrone, pinder

Please Log in or Create an account to join the conversation.

More
02 Oct 2020 02:07 #184528 by txtrone
Same Lenovo E430 BUT with upgraded RAM and new SSD drive. Tuned the BIOS as best I know how. Mint 19.3
LinuxCNC pc tests, version 20200930, started 01.10.2020 18:44:13
********************************************************************************
General info:
LENOVO,  3254CTO
Bios version H0ET98WW (2.58 ), dated 08/18/2016
Chipset: Intel Corporation 7 Series/C216 
Harddisk is non-rotational, ie. SSD
The amount of swap currently used is 0
================================================================================
CPU related data:
model name	: Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz
cpu cores	: 2
stepping	: 9
cache size	: 3072 KB
================================================================================
RAM related data:
Maximum Capacity: 16 GB
Size: 8192 MB
Size: 8192 MB
================================================================================
GPU related data:
product: 3rd Gen Core processor Graphics Controller
configuration: driver=i915 latency=0
================================================================================
Parallel port related data:
number of parallel ports is 0
================================================================================
Os and desktop related data:
PRETTY_NAME="Linux Mint 19.3"

Should the info below not match with your machine,
kindly post the output of pstree here, including a brief
desciption of your desktop environment, window manager and
display manager. Tia
https://forum.linuxcnc.org/18-computer/39370-script-for-automated-testing-of-computer-latency

Desktop environment : lightdm
Windowmanager       : Xfwm4
Displaymanager      : lightdm
================================================================================
Kernel related data:
Kernel 4.19.106-rt46-lcnc
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
Cpu idle driver: intel_idle
================================================================================
Kernel modules data:
Check if pcspkr is loaded: No, which is good!
Check if snd modules are loaded: No, which is good!
================================================================================
Keyboard & Mouse related data:
Number of mice: 2
Mouse 0 : SynPS/2 Synaptics TouchPad
Mouse 1 : TPPS/2 IBM TrackPoint
Mice  attached to USB:
Keyboards attached to USB:
================================================================================
LinuxCNC related data:
LinuxCNC version is: 2.8.0
================================================================================
Preparing for latency test:
Started glxgears No. 1, 2, 3, 4, 5
..........
Started firefox with 'https://vimeo.com/150574260', WID 71303171
================================================================================
Command for latency test is:
su oldlt -c  "latency-histogram --nobase  --sbinsize 1000 --nox 2>&1 | tee -a 1601595852.lat &"

Latency testing loop started Thu Oct  1 18:44:28 CDT 2020, 
should end after Thu Oct  1 20:44:28 CDT 2020

================================================================================
Swap useage: prerun: 0, running: 0

================================================================================
CPU core temperatures:
Core 0: Tmin: 53.0°C, Tmax: 74.0°C, (high = +87.0°C, crit = +105.0°C)
Core 1: Tmin: 56.0°C, Tmax: 80.0°C, (high = +87.0°C, crit = +105.0°C)
================================================================================
last latency data is as follows:.
 7203 secs  servo min:-113.800 uS max: 121.800 uS sdev:   5.900 uS
 7204 secs  servo min:-113.800 uS max: 121.800 uS sdev:   5.900 uS
The following user(s) said Thank You: seuchato

Please Log in or Create an account to join the conversation.

More
02 Oct 2020 10:49 #184550 by seuchato
No problem. We are in no hurry. Support your brother, that's a very good thing to do.
greez
chris
The following user(s) said Thank You: tommylight

Please Log in or Create an account to join the conversation.

More
02 Oct 2020 14:33 #184577 by seuchato

BeagleBrainz wrote: Just a thought.
a) Whilst all this testing is great, the average forum user isn’t going to wade through this many pages.
b) It would really need collating and summarising.
c) Maybe a new topic just listing the machine and latency with a link to the actual test in this thread maybe the go. Keep the summary on the first post, just add an edit for a new test.
....


BeagleBrainz

a) There is off course search function that can shorten the "wading".
b) Agreed, but without c) it will probably end at the same level of insight as in the latency database . To me, the variety of information and quality thereof plus that many hardware details are missing in the wiki database, was the main motivation to write lcnc-hw and ask tommy to open this thread. It will need a high level of discipline and probably a guideline to write up good short comprehensive summaries. Of course if that thread was moderated and entries corrected by say a few authorized people that could be granted. In that case 100% agreed. But who is going to do this?

greez
chris

Please Log in or Create an account to join the conversation.

More
03 Oct 2020 18:43 #184716 by Bari
One of the main issues in getting low latency is getting the BIOS configured properly. BIOSes are also often broken and the settings do not do as they are expected. This is one reason why the same CPU and chipset used on different motherboards will have different latency results.

Things in BIOS that should always be OFF:

Power management

Speed stepping/cpu core speed throttling

Virtualization

Fast Boot

Some of the info on the wiki about getting lower latency is also old. Trying an external graphics card has not helped for PC's made in the past 10 year.

I should rewrite the wiki page about lowering latency and what numbers are actually good vs poor. I should work on this with Mesa to post accurate info since I still see people asking about EPP modes on PCI/e LPT cards when not even using a Mesa FPGA.
The following user(s) said Thank You: seuchato, tommylight

Please Log in or Create an account to join the conversation.

More
04 Oct 2020 01:10 #184767 by seuchato

Bari wrote: One of the main issues in getting low latency is getting the BIOS configured properly. BIOSes are also often broken and the settings do not do as they are expected. This is one reason why the same CPU and chipset used on different motherboards will have different latency results.

Things in BIOS that should always be OFF:

Power management

Speed stepping/cpu core speed throttling

Virtualization


100% agreed...

Bari wrote: Fast Boot


I will retest this, but I have not seen an effect on my pc's. Should my tests bring another result, I will edit this remark correspondingly. However, as we use our pcs for machining, the little extra time for normal boot will not really be harmful.

Bari wrote: Some of the info on the wiki about getting lower latency is also old.

Bari wrote: Trying an external graphics card has not helped for PC's made in the past 10 year.

Not sure about that. I am going to test this on this pc.

Bari wrote: I should rewrite the wiki page about lowering latency and what numbers are actually good vs poor.

That rewrite and especially adding "what numbers are actually good vs poor" would be greatly appreciated

Bari wrote: I should work on this with Mesa to post accurate info since I still see people asking about EPP modes on PCI/e LPT cards when not even using a Mesa FPGA.


Thanks a ton
chris

Please Log in or Create an account to join the conversation.

More
05 Oct 2020 22:47 #185014 by txtrone
Same Lenovo E430. 16GB memory, new SSD, running Buster. I am happy with the results. I am going to run one more test on this rig, with the original HDD in it.
LinuxCNC pc tests, version 20200930, started 05.10.2020 15:37:04
********************************************************************************
General info:
LENOVO,  3254CTO
Bios version H0ET98WW (2.58 ), dated 08/18/2016
Chipset: Intel Corporation 7 Series/C216 
Harddisk is non-rotational, ie. SSD
The amount of swap currently used is 0
================================================================================
CPU related data:
model name	: Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz
cpu cores	: 2
stepping	: 9
cache size	: 3072 KB
================================================================================
RAM related data:
Maximum Capacity: 16 GB
Size: 8192 MB
Size: 8192 MB
================================================================================
GPU related data:
product: 3rd Gen Core processor Graphics Controller
configuration: driver=i915 latency=0
================================================================================
Parallel port related data:
number of parallel ports is 0
================================================================================
Os and desktop related data:
PRETTY_NAME="Debian GNU/Linux 10 (buster)"

Should the info below not match with your machine,
kindly post the output of pstree here, including a brief
desciption of your desktop environment, window manager and
display manager. Tia
https://forum.linuxcnc.org/18-computer/39370-script-for-automated-testing-of-computer-latency

Desktop environment : xfce4
Windowmanager       : Xfwm4
Displaymanager      : lightdm
================================================================================
Kernel related data:
Kernel 4.19.0-10-rt-amd64
GRUB_CMDLINE_LINUX_DEFAULT="quiet"
Cpu idle driver: intel_idle
================================================================================
Kernel modules data:
Check if pcspkr is loaded: No, which is good!
Check if snd modules are loaded: No, which is good!
================================================================================
Keyboard & Mouse related data:
Number of mice: 2
Mouse 0 : SynPS/2 Synaptics TouchPad
Mouse 1 : TPPS/2 IBM TrackPoint
Mice  attached to USB:
Keyboards attached to USB:
================================================================================
LinuxCNC related data:
LinuxCNC version is: 2.8.0
================================================================================
Preparing for latency test:
Started glxgears No. 1, 2, 3, 4, 5
..........
Started firefox with 'https://vimeo.com/150574260', WID 50331651
================================================================================
Command for latency test is:
su oldlt -c  "latency-histogram --nobase  --sbinsize 1000 --nox 2>&1 | tee -a 1601930222.lat &"

Latency testing loop started Mon 05 Oct 2020 03:37:18 PM CDT, 
should end after Mon 05 Oct 2020 05:37:18 PM CDT

================================================================================
Swap useage: prerun: 0, running: 0

================================================================================
CPU core temperatures:
Core 0: Tmin: 54.0°C, Tmax: 71.0°C, (high = +87.0°C, crit = +105.0°C)
Core 1: Tmin: 56.0°C, Tmax: 78.0°C, (high = +87.0°C, crit = +105.0°C)
================================================================================
last latency data is as follows:.
 7202 secs  servo min: -73.900 uS max:  73.700 uS sdev:   5.100 uS
 7203 secs  servo min: -73.900 uS max:  73.700 uS sdev:   5.100 uS

Please Log in or Create an account to join the conversation.

More
06 Oct 2020 04:34 #185033 by txtrone
Same as above, bit with 7200 HDD.
LinuxCNC pc tests, version 20200930, started 05.10.2020 18:30:41
********************************************************************************
General info:
LENOVO,  3254CTO
Bios version H0ET98WW (2.58 ), dated 08/18/2016
Chipset: Intel Corporation 7 Series/C216 
Harddisk is rotational
The amount of swap currently used is 0
================================================================================
CPU related data:
model name	: Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz
cpu cores	: 2
stepping	: 9
cache size	: 3072 KB
================================================================================
RAM related data:
Maximum Capacity: 16 GB
Size: 8192 MB
Size: 8192 MB
================================================================================
GPU related data:
product: 3rd Gen Core processor Graphics Controller
configuration: driver=i915 latency=0
================================================================================
Parallel port related data:
number of parallel ports is 0
================================================================================
Os and desktop related data:
PRETTY_NAME="Debian GNU/Linux 10 (buster)"

Should the info below not match with your machine,
kindly post the output of pstree here, including a brief
desciption of your desktop environment, window manager and
display manager. Tia
https://forum.linuxcnc.org/18-computer/39370-script-for-automated-testing-of-computer-latency

Desktop environment : xfce4
Windowmanager       : Xfwm4
Displaymanager      : lightdm
================================================================================
Kernel related data:
Kernel 4.19.0-10-rt-amd64
GRUB_CMDLINE_LINUX_DEFAULT="quiet"
Cpu idle driver: intel_idle
================================================================================
Kernel modules data:
Check if pcspkr is loaded: No, which is good!
Check if snd modules are loaded: No, which is good!
================================================================================
Keyboard & Mouse related data:
Number of mice: 2
Mouse 0 : SynPS/2 Synaptics TouchPad
Mouse 1 : TPPS/2 IBM TrackPoint
Mice  attached to USB:
Keyboards attached to USB:
================================================================================
LinuxCNC related data:
LinuxCNC version is: 2.8.0
================================================================================
Preparing for latency test:
Started glxgears No. 1, 2, 3, 4, 5
..........
Started firefox with 'https://vimeo.com/150574260', WID 48234499
================================================================================
Command for latency test is:
su oldlt -c  "latency-histogram --nobase  --sbinsize 1000 --nox 2>&1 | tee -a 1601940640.lat &"

Latency testing loop started Mon 05 Oct 2020 06:30:57 PM CDT, 
should end after Mon 05 Oct 2020 08:30:57 PM CDT

================================================================================
Swap useage: prerun: 0, running: 0

================================================================================
CPU core temperatures:
Core 0: Tmin: 54.0°C, Tmax: 68.0°C, (high = +87.0°C, crit = +105.0°C)
Core 1: Tmin: 56.0°C, Tmax: 72.0°C, (high = +87.0°C, crit = +105.0°C)
================================================================================
last latency data is as follows:.
 7201 secs  servo min: -55.400 uS max:  55.100 uS sdev:   3.300 uS
 7202 secs  servo min: -55.400 uS max:  55.100 uS sdev:   3.300 uS

Please Log in or Create an account to join the conversation.

More
06 Oct 2020 07:22 #185038 by seuchato
txtrone

Quite astonishing that the results with hd give better results than with ssd!
I guess, tommylight would agree: really usable.

Greez
Chris

Please Log in or Create an account to join the conversation.

Time to create page: 0.138 seconds
Powered by Kunena Forum