FORUMS
Remove All Ads from XDA

 View Poll Results: What CPU governor do you use?

Ondemand/OndemandX
 
233 Vote(s)
7.05%
Interactive/InteractiveX
 
1,053 Vote(s)
31.87%
Conservative/Ktoonservative
 
121 Vote(s)
3.66%
Performance
 
139 Vote(s)
4.21%
Powersave
 
81 Vote(s)
2.45%
ZZMoove
 
157 Vote(s)
4.75%
Intelliactive/Intellidemand
 
98 Vote(s)
2.97%
Hyper
 
23 Vote(s)
0.70%
PegasusQ
 
45 Vote(s)
1.36%
Impulse
 
122 Vote(s)
3.69%
Smartmax/Performance May Cry
 
62 Vote(s)
1.88%
ElementalX
 
193 Vote(s)
5.84%
Alucard
 
225 Vote(s)
6.81%
Yankactive/YanksusQ/Yankdemand
 
26 Vote(s)
0.79%
Darkness/Nightmare/Lightning
 
92 Vote(s)
2.78%
Blu_active
 
247 Vote(s)
7.48%
SmartassV2/SmartassH3
 
67 Vote(s)
2.03%
Other
 
320 Vote(s)
9.69%

[REF][GUIDE]Most up to date guide on CPU governors, I/O schedulers and more!

4,021 posts
Thanks Meter: 8,377
 
By Saber, Recognized Contributor on 8th March 2015, 05:53 AM
Post Reply Email Thread
24th June 2018, 01:44 PM |#891  
runeupo's Avatar
Senior Member
Thanks Meter: 148
 
More
Quote:
Originally Posted by Cichy26

Can you tell how to tweak interactive for better battery life?

Use 'search plus' before asking, tweak interactive..


https://forum.xda-developers.com/nex...ernor-t3290605
The Following 3 Users Say Thank You to runeupo For This Useful Post: [ View ] Gift runeupo Ad-Free
15th July 2018, 12:47 AM |#892  
Infy_AsiX's Avatar
Senior Member
Flag Brisbane
Thanks Meter: 381
 
More
Quote:
Originally Posted by koyo

Hey. I looked up things about Android and why it's UI lags so much. I once stumbled upon a topic that said this was the fault of the CPU governor, as it was the frequency jumps that caused this and not anything inherent. Said poster then proceeded to provide profile GPU pics of stuttering between a locked CPU frequency and one that uses the governor, and it indeed displayed less dropped frames (iirc, when the graph shoots past the green line, it means frames have been dropped). Is this just some placebo effect? Should I be using the performance governor if I want smooth UI since it locks the frequency?

Never properly tested developer settings Profile GPU rendering until recently. I also saw better UI responsiveness and close to no stutter on performance governor. Thus figured out scrolling stutter inherent on S820 was CPU governor related.

I'd no success in tweaking interactive or ondemand higher. What baffled is my Z3C S801 has similarly tuned aggressive interactive and maintains responsiveness. But I had TripNDroid @TripNRaVeR governor ready on the kernel. Setting it's down_sample_time 150000 from 30000 almost matches the UI responsiveness of performance gov, no significant scroll stuttering now except immediately after switching apps or fresh loading an intensive app like Play Store. No touch boost required, which means long several second extended flick scrolling doesn't lag out and that the gov is efficient at handling demand appropriately without relying on arbitrary touch boosting. Note, I've also ROM thermal throttling disabled.

Don't know why I couldn't tune Interactive or Ondemand to perform enough, however i'm no expert. It does lead me to think Snapdragon was UI compromised hard tuned for less power due to battery life and heat limitations of the spec and OS platform. Yet the right governor can tune things high enough for UI responsiveness.

update edit: I also just realised tripndroid suffers stuttering too without adrenoboost higher than low setting too. Only recently discovered adrenoboost, it scales really well for performance, turns out eliminating stuttering is high performance cpu and gpu
The Following 2 Users Say Thank You to Infy_AsiX For This Useful Post: [ View ] Gift Infy_AsiX Ad-Free
18th July 2018, 08:20 AM |#893  
andrea0807's Avatar
Senior Member
Flag Perugia
Thanks Meter: 22
 
More
Hi guys, which of these governors has a good battery life without affecting the performance too much? lionfish could be okay?Click image for larger version

Name:	Screenshot_Kernel_Adiutor_20180718-091705.jpeg
Views:	1970
Size:	23.3 KB
ID:	4552353

Inviato dal mio LEX820 utilizzando Tapatalk
18th July 2018, 10:53 AM |#894  
nexus_vl's Avatar
Senior Member
Flag Mol
Thanks Meter: 292
 
Donate to Me
More
Quote:
Originally Posted by andrea0807

Hi guys, which of these governors has a good battery life without affecting the performance too much? lionfish could be okay?Attachment 4552353

Inviato dal mio LEX820 utilizzando Tapatalk

It looks like you're using an EAS based kernel, then you should definitely use schedutil for best battery
The Following User Says Thank You to nexus_vl For This Useful Post: [ View ] Gift nexus_vl Ad-Free
19th August 2018, 05:13 AM |#896  
gedega10's Avatar
Member
Thanks Meter: 3
 
More
Quote:
Originally Posted by Saber

TCP algorithms guide


What are TCP algorithms?
Congestion control strategies (or algorithms) are used by TCP, the data transmission protocol used by many Internet applications. The main goal of a TCP algorithm is to avoid sending more data than the network is capable of transmitting, that is, to avoid causing network congestion. Different algorithms respond differently to network loads, but they are all based on the same principle of avoiding network congestion.

Things to look out for in TCP algorithms include (but not exclusively):
- Download/Upload speeds - The higher the number, the better
- Latency - The lower the number, the better


TCP Algorithm Descriptions


Tahoe:



Reno:


Vegas:



Hybla:


Cubic:


Westwood/Westwood+:



Low Priority (LP):



Binary Increase Congestion control (BIC):



Scalable:



Hamilton TCP (HTCP):


Veno:



Illinois:



High speed (HSTCP):



Yeah-TCP:



CDG


Benchmark

Results:

Latency - Download - Upload


cubic:

1st run: 15ms - 10,75Mbps - 7,82Mbps
2nd run: 14ms - 10,84Mbps - 8,06Mbps

reno:
1st run: 13ms - 15,51Mbps - 6,73Mbps
2nd run: 13ms - 14,73Mbps - 8,51Mbps

bic:
1st run: 12ms - 10,38Mbps - 8,61Mbps
2nd run: 13ms - 10,78Mbps - 8,62Mbps

westwood:
1st run: 11ms - 17,65Mbps - 8,30Mbps
2nd run: 13ms - 13,28Mbps - 8,29Mbps

highspeed:
1st run: 13ms - 10,76Mbps - 7,94Mbps
2nd run: 16ms - 14,42Mbps - 8,52Mbps

hybla:
1st run: 14ms - 11,19Mbps - 7,44Mbps
2nd run: 14ms - 13,47Mbps - 7,56Mbps

htcp:
1st run: 14ms - 13,24Mbps - 7,03Mbps
2nd run: 15ms - 10,85Mbps - 8,00Mbps

vegas:
1st run: 14ms - 8,49Mbps - 6,62Mbps
2nd run: 14ms - 12,00Mbps - 7,07Mbps

veno:

1st run: 13ms - 9,58Mbps - 8,13Mbps
2nd run: 13ms - 8,50Mbps - 7,64Mbps

scalable:
1st run: 18ms - 12,01Mbps - 8,73Mbps
2nd run: 14ms - 13,96Mbps - 8,23Mbps

lp:
1st run: 14ms - 14,90Mbps - 8,68Mbps
2nd run: 14ms - 13,44Mbps - 8,72Mbps

yeah:
1st run: 14ms - 13,37Mbps - 8,28Mbps
2nd run: 17ms - 13,89Mbps - 8,14Mbps

illinois:
1st run: 13ms - 12,93Mbps - 8,24Mbps
2nd run: 16ms - 13,97Mbps - 6,46Mbps

Recommendations:



For speed:
- Westwood - Best
- Highspeed
- LP

For stability:

- Cubic - Best
- Reno

For high latency networks:

- Westwood - Best
- BIC

For general usage:
- Cubic - Best
- Westwood - Best
- Reno


Conclusion:

The only TCP algorithms I would recommend are Cubic or Westwood as they are the most stable and efficient for mobile devices. Real world usage would show little difference when comparing between algorithms, so everyone's experience will vary! There are myths that changing algorithms will affect battery life, this is not true!!!!!

Also, don't be surprised if you end up using Cubic or Westwood!



Hi Saber, I found some kernels have new TCP's, care to research and teach us? Thanks much!
19th August 2018, 10:18 AM |#897  
Saber's Avatar
OP Recognized Contributor
Flag Melbourne
Thanks Meter: 8,377
 
Donate to Me
More
Quote:
Originally Posted by gedega10

Hi Saber, I found some kernels have new TCP's, care to research and teach us? Thanks much!

Sure, when I have time, I'll do some research
The Following 3 Users Say Thank You to Saber For This Useful Post: [ View ] Gift Saber Ad-Free
20th August 2018, 09:02 AM |#898  
Member
Thanks Meter: 7
 
More
what would happen if i apply more than one hotplug?
20th August 2018, 11:28 AM |#899  
marcdw's Avatar
Senior Member
Flag Baldwin Park
Thanks Meter: 2,906
 
More
Quote:
Originally Posted by damarwardoyo

what would happen if i apply more than one hotplug?

Doesn't matter. You're to use only one. Period.
If you want to enable more than one for some strange reason, go ahead. Report back.


Moto G5S Plus XT1806, balanceOS 2.0, MultiROM, XDA Legacy
The Following 2 Users Say Thank You to marcdw For This Useful Post: [ View ] Gift marcdw Ad-Free
21st August 2018, 12:46 AM |#900  
Saber's Avatar
OP Recognized Contributor
Flag Melbourne
Thanks Meter: 8,377
 
Donate to Me
More
Quote:
Originally Posted by damarwardoyo

what would happen if i apply more than one hotplug?

Generally only one hotplug driver should be used. When two or more hotplugs run at once, unexpected behavior can occur (i.e poor battery life, poor performance, stability issues, etc)
21st August 2018, 08:17 AM |#901  
zr_hazem's Avatar
Senior Member
Thanks Meter: 53
 
More
@Saber can you please provide us with an explanation or documentation to schedutil tunables such as sched_nr_migrate, sched_latency_ns. There are many more and I'd like to understand what they do exactly.
Post Reply Subscribe to Thread

Tags
cpu governor, graphs, gsstudios, guide, io scheduler

Guest Quick Reply (no urls or BBcode)
Message:
Previous Thread Next Thread
Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes