1. International Symposium on Microarchitecture. New York, NY.
INTO THE WILD Studying Real User Activity Patterns to Guide Power Optimizations for Mobile Architectures
Alex Shye
Ben Scholbrock
Gokhan Memik
Northwestern University, EECS
Empathic Systems Project
12/14/2009
International Symposium on Microarchitecture. New York, NY.

2. The rise of mobile architectures
ENIAC
1946
Apple II
1977
Media players, PDAs, smartphones, netbooks
Today
IBM System/360
1964
IBM Thinkpad 700
1992
1940
1950
1960
1970
1980
1990
2000
2010
Trends
Implications
 Form factor
 Total energy store
Power optimization is important
 “Personal” nature
The user drives execution
12/14/2009
International Symposium on Microarchitecture. New York, NY.

3. International Symposium on Microarchitecture. New York, NY.
Summary
Observation: For mobile architectures, the user is the workload
Claim: Computer architects should study user activity to:
(1) Characterize power consumption
(2) Guide the development of power optimizations
12/14/2009
International Symposium on Microarchitecture. New York, NY.

4. International Symposium on Microarchitecture. New York, NY.
Summary
Observation: For mobile architectures, the user is the workload
Claim: Computer architects should study user activity to:
(1) Characterize power consumption
Log the activity of real users using the Android G1
Develop an accurate power model for the Android G1
Power consumption varies across users
The screen and CPU consume the most power
(2) Guide the development of power optimizations
Screen time dominated by long screen intervals
Develop optimizations leveraging change blindness
Save ~10% total system power with minimal change to user satisfaction
12/14/2009
International Symposium on Microarchitecture. New York, NY.

5. Target Mobile Architecture
We study the HTC Dream mobile platform (Android G1)
First platform based upon Android OS
2 phones, identical hardware:
Google Android Developer Phone 1 (ADP1)
T-Mobile G1 – commercialized version of the G1
[Google ADP1]
How do we proceed with power optimizations?
Which components consume the most power?
12/14/2009
International Symposium on Microarchitecture. New York, NY.

6. Logging Real User Activity
NU JamLogger
Runs on any Android G1
Logs system activity
CPU utilization, Wifi traffic, SD card traffic, screen usage, etc.
Periodically sends logs to our server
Lightweight
< 5% CPU overhead during active phone use
Logs compressed with gzip before upload
Getting Users
Released NU JamLogger on Android Market
Advertised at Northwestern University, University of Michigan, and online (Slashdot, online Android forums, etc)
20 users, ~250 days of cumulative user activity
Each user has over 1 weeks worth of logs
12/14/2009
International Symposium on Microarchitecture. New York, NY.

7. Power Estimation Model
HW Unit
Parameter
CPU
hi_cpu_util
med_cpu_util
Screen
screen_on
screen_brightness
Call
call_ringing
call_off_hook
EDGE
edge_has_traffic
edge_traffic
Wifi
wifi_on
wifi_has_traffic
wifi_traffic
SD Card
sdcard_traffic
DSP
music_on
System
system_on
Idle
idle
Jam Logs …
49446 : CPU_Utilization
49491 : Load_Avg
50343 : Screen_Off
50557 : Wifi_Traffic 10 0
Power Measurements
Instrumented
Battery via Current Clamp
Time-based Traces
1 sec. samples
w/ parameters
Linear
Regression
Power Model
Phone does not provide run-time power measurements
12/14/2009
International Symposium on Microarchitecture. New York, NY.

8. Accuracy of Power Model
Hardware Specific Runs
Usage Scenarios
Built power model with one ADP1, validated on a separate ADP1
6.6% median error per sample
< .1% error summed across all samples (total energy)
12/14/2009
International Symposium on Microarchitecture. New York, NY.

9. International Symposium on Microarchitecture. New York, NY.
Using the Power Model
CPU
Wifi
DSP
Screen
One example test run
Estimated power closely tracks real-time measurements
Can sum parameters to derive a power breakdown
12/14/2009
International Symposium on Microarchitecture. New York, NY.

10. Power Breakdown (20 Users)
There exists a significant variation in users
Idle is very important, consuming 49.3% of total energy
12/14/2009
International Symposium on Microarchitecture. New York, NY.

11. Power Breakdown (20 Users w/o Idle)
Ignoring Idle time, the screen and CPU are most power hungry
Screen: 35.5% (19.2% from brightness)
CPU: 12.7%
12/14/2009
International Symposium on Microarchitecture. New York, NY.

12. Studying User Activity for Optimization
Screen activity is a good indicator of usage activity
We study screen intervals, periods of time the screen is on
Observation: The total screen time is dominated by a small number of long screen intervals
Screen intervals of 100+ seconds account for ~70% of screen time
We develop a power optimization strategy that leverages change blindness to target the screen and CPU
12/14/2009
International Symposium on Microarchitecture. New York, NY.

13. International Symposium on Microarchitecture. New York, NY.
Change Blindness
The inability to distinguish changes in a stimulus
[Simons 99]
12/14/2009
International Symposium on Microarchitecture. New York, NY.

14. International Symposium on Microarchitecture. New York, NY.
Ramp Optimizations
Can change blindness be applied for power optimization?
We target the screen and CPU for optimization:
Screen Ramp: slowly dim brightness
Decreases brightness by 7 units (max 255) until 60% of starting brightness
CPU Ramp: slowly decrease effective frequency
Reaches 70% of maximum frequency in 40 seconds
Implemented by tuning ondemand DFS governor
We compare to Drop optimizations
Screen Drop: drop to 60% brightness after 40 seconds
CPU Drop: drop to 70% frequency after 40 seconds
12/14/2009
International Symposium on Microarchitecture. New York, NY.

15. International Symposium on Microarchitecture. New York, NY.
Power Savings
Simulate optimizations on real user traces
CPU Ramp saves 4.9% system power (22.8% of CPU power)
Screen Ramp saves 5.7% system power (19.1% of screen power)
Ramp optimizations outperform Drop optimizations
12/14/2009
International Symposium on Microarchitecture. New York, NY.

16. Impact on User Experience
User study design
Run four optimizations and a control run (in random order), on:
Web browsing: Browse Wikipedia
Game: BreakTheBricks game
Video: Watching video with PlayVideo
Ask for verbal satisfaction rating from 1 (low) – 5 (high)
At end of study, debrief user, and discuss acceptance of optimizations
Results
No significant difference in user satisfaction when compared to control (using paired t-test) except for small changes in:
All optimizations with the Game
Screen Drop with Video
Users mostly rated based upon feeling of jumpiness/jitter
Works well for screen; most users couldn’t tell screen was changing
15 users would use some of optimizations, 4 would not, 1 apathetic
More details in paper
12/14/2009
International Symposium on Microarchitecture. New York, NY.

17. International Symposium on Microarchitecture. New York, NY.
Conclusion
We study real users on real devices in real environments
Develop and user power model to characterize power consumption
Show that power consumption does vary across users
When active, the screen and CPU consume the most power
Present an example of a user-activity-driven power optimization
Develop ramp optimizations targeting the screen and CPU
Save 5.7% and 4.9% of total system power for the screen and CPU, respectively
Ramp optimizations well-accepted by users, especially for the screen
Computer architects should study real user activity to:
(1) Characterize power consumption
(2) Guide the development of power optimizations
12/14/2009
International Symposium on Microarchitecture. New York, NY.

18. International Symposium on Microarchitecture. New York, NY.
Thank You!
Questions?
Alex Shye
ESP: Empathic Systems Project
NU JamLogger: Available on Android Market
12/14/2009
International Symposium on Microarchitecture. New York, NY.