Breaking News
News & Analysis

Big.little test results show promise, ARM reports

A7 matches, A15 beats A9 core
11/5/2012 07:07 AM EST
8 comments
NO RATINGS
< Previous Page 2 / 6 Next >
More Related Links
View Comments: Oldest First | Newest First | Threaded View
wsw1982
User Rank
Rookie
re: Big.little test results show promise, ARM reports
wsw1982   11/5/2012 10:15:12 AM
NO RATINGS
That's different form what observed from Samsung Chromebook and Nexus 10. The Exynos 5 is power hungry.

rick merritt
User Rank
Author
re: Big.little test results show promise, ARM reports
rick merritt   11/5/2012 3:15:19 PM
NO RATINGS
The Samsung Exynos 5 uses two A15 cores which are pretty power hungry. In the A7/A15 big.little approach, the A7 does most of the work and the A15 just powers up a minority of the time to blast through intense jobs. That should cut power by 150% or more, ARM estimates. Now I wonder, will Intel adopt this concept in mobile?

glennthera
User Rank
Rookie
re: Big.little test results show promise, ARM reports
glennthera   11/5/2012 4:53:28 PM
NO RATINGS
Thanks 4 d update Rick, been waiting on some news frm ARM n Samsung 4 a long, long time! big.Little is d future of processors, not jst SOC designs!

danny1024
User Rank
Rookie
re: Big.little test results show promise, ARM reports
danny1024   11/5/2012 8:15:23 PM
NO RATINGS
There have been many research projects that have looked at combining Atom and "big cores" e.g. Ivy Bridge and Haswell as well as ARM processors to handle low-level OS functions and some applications while the bigger cores are sleeping. In fact, I think Dell had a laptop that had both "ARM & Intel Inside" to do roughly the same thing. The A15 is a hot, complex chip that's not particularly performant for the power nor does it yield particularly well. That A7 companion chip and the entire Big.Little concept is a tacit admission of that regrettable reality.

rick merritt
User Rank
Author
re: Big.little test results show promise, ARM reports
rick merritt   11/5/2012 9:06:04 PM
NO RATINGS
Can any share some performance per watt data they have seen on an A15?

danny1024
User Rank
Rookie
re: Big.little test results show promise, ARM reports
danny1024   11/6/2012 4:49:38 PM
NO RATINGS
I'll ask around but I can tell you that many of the A15 licensees are greatly disappointed with its perf/watt and associated manufacturability issues.

help.fulguy
User Rank
Manager
re: Big.little test results show promise, ARM reports
help.fulguy   11/10/2012 7:15:19 PM
NO RATINGS
@ Rick, a Dual core A15 @ 2GHz in our lab is consuming 4W. Mind you, this is just the CPU Cores. When we add the Graphics, Video and IOs, you are looking at around 6-7W. ARM is in big trouble. This big.Little is just a hog-wash. Unless we see this from a 3rd party, I cant believe 50% improvement. ARM CEO can say Power this and power that, but his ARM Cores are getting worse by day.

zeewee
User Rank
Rookie
re: Big.little test results show promise, ARM reports
zeewee   11/20/2012 4:37:05 AM
NO RATINGS
I think it's more efficient to improve its power saving mode such as Intel's CPU's C-state.

Flash Poll
Radio
LATEST ARCHIVED BROADCAST
Join our online Radio Show on Friday 11th July starting at 2:00pm Eastern, when EETimes editor of all things fun and interesting, Max Maxfield, and embedded systems expert, Jack Ganssle, will debate as to just what is, and is not, and embedded system.
Like Us on Facebook

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
EE Times on Twitter
EE Times Twitter Feed
Top Comments of the Week