Latency of ext. mouse & ext. display with USB 3 Port Replicator and Celeron Chromebooks
Reported by
michael....@hp.com,
Dec 22 2016
|
|||||
Issue description
Chrome Version : 54.0.2840.101 and 55
OS Version: 10.0
URLs (if applicable) :
Other browsers tested:
Add OK or FAIL after other browsers where you have tested this issue:
Safari 5:--
Firefox 4.x:--
IE 7/8/9:--
What steps will reproduce the problem?
Lag / latency of the mouse (or display complete itself) is very small on the extern display (24” 1920x1080) which is connected through a USB3 port replicator (e.g. Fujitsu PR08). The feedback is slow while hovering over active menus, or moving a complete small window.
This occurs with HP Chromebooks with Intel N2840 and N2940 CPUs (Chrome FW 54.0.2840.101), as well with an Acer Chromebook (CB3-131-C1CA) N2840 CPU (Chrome FW 4.0.2840.101).
This occurs as well with HP USB 3 Docking pr3005, this is same DisplayLink technology as Fujitsu.
While moving the mouse over the external display the system monitor shows a load of about 90%.
No difference is mouse is connected at the PortReplicator or direct (or usage of the internal mouse).
No difference if display is connected through DP or DVI to the PortReplicator.
No difference which Chromebook USB ports are used.
.
Using the internal chromebook display the mouse movement, reaction and feedback from display is perfect.
On the CB internal monitor the system monitor shows was a little bit less about 70-80%.
Only moving the mouse over the external display COG sysmonitor shows about 75%, on the internal display it shows about 25% CPU usage.
Using a HP Chromebook 13 with Pentium 4405Y close to perfect, only very yery small latency, acceptable by user
Using the Google Pixel or a standard Intel Notebook, no issue as well.
While moving the mouse over the external display the system monitor shows a load of about 50%.
On the internal monitor it was a little bit less.
For all tested system: using the internal HDMI of the HP or Acer Chromebook, no issue, perfect “feedback” from mouse and display.
Idea: Could be there an optimized driver for the PortReplicator?
What is the expected result?
No latency of the mouse & external Display whil using a USB PortReplicator.
What happens instead of that?
Please provide any additional information below. Attach a screenshot if
possible.
,
Dec 27 2016
,
Jan 20 2017
Add displaylink to investigate
,
Jan 21 2017
Note that the systems which are problematic are using old and slow small core Bay Trail based chipsets (e.g. N2840). If the DisplayLink adapter requires video compression, it is likely to cause a heavy load on such systems, which may also vary based on activity on the external USB display (e.g. mouse motion). Additionally these older models had both USB 2.0 and USB 3.0 ports, using the USB 2.0 port may be causing additional compression to be necessary for the display. It may be worth trying the same USB display adapter using a Windows based Bay Trail system to see if similar performance problems are present to try to give a baseline, this should give us some idea if this performance level is expected behavior on such hardware configurations.
,
Jan 21 2017
I don't think we can do much here. By design for displaylink the CPU has to be involved, if the CPU is slow, then that's pretty much it.
,
Jan 24 2017
Are there possible changes on the OS side that could possibly mitigate this? I had a feeling this was a hardware limitation from the beginning as this issue shows on all platforms with this chipset. However since this is such a large presales case would like to try anything to help the customer. I am willing to ship unit/monitor/dock if that is necessary. Please let me know if you are willing or if this is the final statement. Thanks Guys!
,
Jan 25 2017
I don't think there is much we could do in this case if the DisplayLink codec has to use the CPU, I don't think it is feasible for it to use hardware acceleration (GPU or otherwise), but I would defer to the DisplayLink folks on that. If a Windows BayTrail system exhibits the same behavior, we are likely stuck with it for these old systems. If Windows can do better on the same BayTrail hardware that would imply there is some chance we could improve the software, but even then we have no guarantee as it may not be easily optimized.
,
Feb 12 2018
Issue has not been modified or commented on in the last 365 days, please re-open or file a new bug if this is still an issue. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot |
|||||
►
Sign in to add a comment |
|||||
Comment 1 by michael....@hp.com
, Dec 22 2016