carlnues
/
paper-KeepItSimple
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

edits (cleanup)

master
carlnues@buffalo.edu 2023-08-26 07:44:23 -04:00
parent 0cdd006baa
commit cdba641add
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
sections/wasted.tex
View File

@ -133,7 +133,7 @@ However, even at the CPU's maximum frequency, more work is created than the CPU
\label{fig:u_micro_fb}
\end{figure}
\Cref{fig:u_micro_fb} shows power consumption for the Facebook workload, padded with idle time to a fixed 40s period.
\Cref{fig:u_micro_fb} shows power consumption for the Facebook workload, padded with idle time to a fixed 60s period.
Operating the CPU at maximum frequency imposes an energy overhead of approximately .6 mAh compared to operating at $\fenergy \approx 70\%$ of its maximum.
This represents about $\frac{1}{1700}$ of the typical Pixel 2's maximum battery capacity.
@ -202,8 +202,8 @@ These observations, coupled with already extant adoption of the \texttt{prio\_hi
\systemname leverages boost requests provided through userspace.
Recall that the \texttt{/dev/stune} API can be used to assign each task a \texttt{boost} value.
Instead of scaling usage history (like \schedutil), \systemname instead simply treats a boost value as a direct request for CPU performance.
\systemname selects the highest frequency of any scheduled task, with $\fenergy$ as a default lower-bound.
Instead of scaling usage history (like \schedutil), \systemname instead simply treats a boost value as a direct request for increased CPU performance.
\systemname selects the highest frequency of any task, with $\fenergy$ as a default lower-bound.
If no tasks are pending, it idles the CPU.
%%% A2