Breaking News
News & Analysis

Samsung Edges Q'comm in New Android Test

6/18/2014 05:00 AM EDT
6 comments
NO RATINGS
More Related Links
View Comments: Newest First | Oldest First | Threaded View
Anand.Yaligar
User Rank
Author
Re: Calling all benchmark gurus
Anand.Yaligar   6/30/2014 12:46:30 AM
NO RATINGS
I think technology node (28nm Vs ??) should also be considered for bench marking, Typically makes 20-30% performance/power/area differences.

@asic_pal, I agree with you. I am curious to know on what process Samsung Exynos is built on ?

Anand.Yaligar
User Rank
Author
Re : Samsung Edges Q'comm in New Android Test
Anand.Yaligar   6/30/2014 12:44:02 AM
NO RATINGS
The Samsung Exynos has nudged ahead of Qualcomm's Snapdragon, according to initial tests of the latest SoCs with a new Android benchmark released today.


Interesting to know that Samsung Exynos has nudged ahead of Qualcomm's Snapdragon. I am curious to know if these are the final numbers or initial numbers because it is mentioned as initial tests in the article ?


Denis.Giri
User Rank
Author
Re: Calling all benchmark gurus
Denis.Giri   6/19/2014 2:42:58 AM
NO RATINGS
Based on the screenshot, "device score" looks like an overall score that actually includes 3D, and if it is to be of any use, this overall score should be taken by itself. Therefore, the older (?) snapdragon is better than the exynos, which is better than the newer snapdragon.

This basically means that either qualcomm are incapable of improving upon themselves, or that this benchmark's overall score is useless: I'm betting on the overall score being useless. better look each individual test score and get an idea of what type of computation each of these devices is better at than rely on an unproven overall score

So samsung edging out qualcomm? not proven yet.

Kresearch
User Rank
Author
Re: Calling all benchmark gurus
Kresearch   6/18/2014 5:26:58 PM
NO RATINGS
Apple to Orange comparison.

asic_pal
User Rank
Author
Re: Calling all benchmark gurus
asic_pal   6/18/2014 2:00:31 PM
NO RATINGS
I think technology node (28nm Vs ??) should also be considered for bench marking, Typically makes 20-30% performance/power/area differences.

rick merritt
User Rank
Author
Calling all benchmark gurus
rick merritt   6/18/2014 10:46:53 AM
NO RATINGS
What do you think of this one?

Most Recent Comments
michigan0
 
SteveHarris0
 
realjjj
 
SteveHarris0
 
SteveHarris0
 
VicVat
 
Les_Slater
 
SSDWEM
 
witeken
Most Recent Messages
9/25/2016
4:48:30 PM
michigan0 Sang Kim First, 28nm bulk is in volume manufacturing for several years by the major semiconductor companies but not 28nm FDSOI today yet. Why not? Simply because unlike 28nm bulk the LDD(Lightly Doped Drain) to minimize hot carrier generation can't be implemented in 28nm FDSOI. Furthermore, hot carrier reliability becomes worse with scaling, That is the major reason why 28nm FDSOI is not manufacturable today and will not be. Second, how can you suppress the leakage currents from such ultra short 7nm due to the short channel effects? How thin SOI thickness is required to prevent punch-through of un-dopped 7nm FDSOI? Possibly less than 4nm. Depositing such an ultra thin film less then 4nm filum uniformly and reliably over 12" wafers at the manufacturing line is extremely difficult or not even manufacturable. If not manufacturable, the 7nm FDSOI debate is over!Third, what happens when hot carriers are generated near the drain at normal operation of 7nm FDSOI? Electrons go to the positively biased drain with no harm but where the holes to go? The holes can't go to the substrate because of the thin BOX layer. Some holes may become trapped at the BOX layer causing Vt shift. However, the vast majority of holes drift through the the un-dopped SOI channel toward the N+Source,...

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed