The Intel Atom Thread

[liahos1]

So 8.68h with a 31WH battery with a chip that is supposed
to be less power hungry than the highly hyped 3770 that was
supposed to eat 2.5W with all four cores running at 2.2-2.3GHz ,
wich should translate to 0.5W with a trivial task like browsing , yet
it seems that s it s several time this amount given said battery life...

What happened.???..And still no official TDP figure , only
sloppy SDPs...

im sorry you're so dissapointed abwx. you can go buy your nexus 7 with 1/3 the performance for better battery life or you can pick from a wide assortment of yet to be announced/designed/delivered amd tablets and get 3 hours of battery life. or you can get an overpriced ipad and have fun on safari! as a consumer you have so many choices!!!

 

[monstercameron]

im sorry you're so dissapointed abwx. you can go buy your nexus 7 with 1/3 the CPU performance BUT SIMILAR GPU PERFORMANCE AND better battery life or you can pick from a wide assortment of yet to be announced/designed/delivered amd tablets and get 3 hours of battery life. or you can get an overpriced ipad and have fun on safari! as a consumer you have so many choices!!!

FTFY, the disappointment I gather is in the marketing buzz around the uber low power draw of the cpu...I for one think the package is a great bargain!

 

[Khato]

FTFY, the disappointment I gather is in the marketing buzz around the uber low power draw of the cpu...

Did I miss something in the review that shows Baytrail to not have excellent power consumption? Because all I saw was a single battery life test which should only be impacted by the SoC if there's something wrong with either it or the OS. You have to remember that the review is of the Asus Transformer Book T100, not Baytrail, and the battery life figures reflect that. The only reason the remainder of the figures might be interesting is to see whether or not Baytrail has any power consumption abnormalities like Haswell with video playback.

 

[SiliconWars]

I think the point is it's showing quite high power draw while simple browsing, system power draw is over 3.5W.

This is one of the problems with budget hardware though, as they can't use the top end components in the likes of the Haswell stuff. In the end, even if Bay Trail has 50% better power consumption it'll be mostly lost when everything is totalled.

I do like the look of this Transformer Book though even without truly amazing power draw, it appears to be aggressively priced.

 

[monstercameron]

Did I miss something in the review that shows Baytrail to not have excellent power consumption? Because all I saw was a single battery life test which should only be impacted by the SoC if there's something wrong with either it or the OS. You have to remember that the review is of the Asus Transformer Book T100, not Baytrail, and the battery life figures reflect that. The only reason the remainder of the figures might be interesting is to see whether or not Baytrail has any power consumption abnormalities like Haswell with video playback.

I should have been more specific, I don't have any qualms with this product or battery life, I was just speculating as to why abwx thought it a bit disappointing.

 

[Abwx]

a wide assortment of yet to be announced/designed/delivered amd tablets and get 3 hours of battery life.

This is deliberate viral marketing based on completely
made up numbers but i guess that some people
dont bother witrh accuracy , but let s check by how
much you are making up your number.

AMD Temash A4-1200

Windows 8
11.6" , not the 10"1 of the T100...

IPS 1366 x 768

8gig memory , far from the 2gig of the T100..

128 Go mass storage.

34.78 Whr battery

9 hours autonomy

http://www.minimachines.net/actu/msi-w20-tablette-11-6-pouces-amd-temash-windows-8-9990

Number is largely comparable to Z3740 given
the size of the battery and the bigger screen ,
so in a case it s a wonder and in another comparable
case it s crap...

 

[Khato]

In the end, even if Bay Trail has 50% better power consumption it'll be mostly lost when everything is totalled.

Definitely true, at least when it comes to light usage models. Sure it'll still make a minor difference when all else is equal, but the quality of other components is going to play a larger part when simply web browsing. Not to mention OS makes a marked difference as you drive power down - hopefully we'll get a Surface 2 review on Tuesday for a good comparison between Baytrail-T and Tegra 4 running their respective versions of windows.

 

[frozentundra123456]

Early number are according to Anand "not stellar" despite :

only 2GB LPDDR3

10.1-inch IPS 1366 x 768

32GB/64GB eMMC + microSD card ,

How much more comsumption with a regular HDD
and a slightly bigger screen , not even at 1080p.??..

Nice selective quote. Here is the entire evaluation of power consumption:

"Battery life looks decent at just over 8.5 hours on a single charge. In practice I had no complaints about battery life while using the device. It feels more like a tablet in that regard and less like a notebook, which is a good thing. Once again we’re seeing ASUS redefining what we’ve come to expect from an entry level notebook PC here. Even compared to Chromebooks we see the T100 do extremely well. I’m curious to get a better feel for how Bay Trail performs in the battery life department, which I’ll be doing over the coming days. So far the results look good but not quite stellar if you compare it to traditional Android/iOS tablets."

 

[Abwx]

Nice selective quote. Here is the entire evaluation of power consumption:

"Battery life looks decent at just over 8.5 hours on a single charge. In practice I had no complaints about battery life while using the device. It feels more like a tablet in that regard and less like a notebook, which is a good thing. Once again we’re seeing ASUS redefining what we’ve come to expect from an entry level notebook PC here. Even compared to Chromebooks we see the T100 do extremely well. I’m curious to get a better feel for how Bay Trail performs in the battery life department, which I’ll be doing over the coming days. So far the results look good but not quite stellar if you compare it to traditional Android/iOS tablets."

If browsing , wich will hardly use more than a core lightly ,
get 8.68h battery life we can assume without much risk
to be significantly off that video playing will hardly last
more than 5 hours or so , i m even sure that Anand has
the numbers but didnt publish them for some reason , heck ,
despite the wonderfull numbers doctored by Intel he didnt
seem to give as much care for what was announced
as the great advantage of this new chip and even relegated
this spzecification to negligible quantity benchmarks wises..

There s no smoke without fire , that s for sure , and all
evidences point to average battery life , far , very far
from what was ultra hyped.

 

[drikkie]

This is deliberate viral marketing based on completely
made up numbers but i guess that some people
dont bother witrh accuracy , but let s check by how
much you are making up your number.

AMD Temash A4-1200

Windows 8
11.6" , not the 10"1 of the T100...

IPS 1366 x 768

8gig memory , far from the 2gig of the T100..

128 Go mass storage.

34.78 Whr battery

9 hours autonomy

http://www.minimachines.net/actu/msi-w20-tablette-11-6-pouces-amd-temash-windows-8-9990

Number is largely comparable to Z3740 given
the size of the battery and the bigger screen ,
so in a case it s a wonder and in another comparable
case it s crap...

You should translate the review better, it has also 2GB of ram, the "Windows 8" quote just before it, threw you off probably.

MSI W20 is a Windows 8 11.6 inch diagonal display on an IPS 1366 x 768 tablet. It will carry a Temash AMD A4-1200, a chip chipset Radeon HD 8180 should offer some comfort in Windows 8. 2 GB of RAM and 128 GB of storage complete this basic equipment.

 

[Abwx]

You should translate the review better, it has also 2GB of ram, the "Windows 8" quote just before it, threw you off probably.

Neverless , the demonstration still stand overhelmingly.

 

[Sweepr]

im sorry you're so dissapointed abwx. you can go buy your nexus 7 with 1/3 the performance for better battery life or you can pick from a wide assortment of yet to be announced/designed/delivered amd tablets and get 3 hours of battery life. or you can get an overpriced ipad and have fun on safari! as a consumer you have so many choices!!!

Worse battery life + total crap CPU performance is what I'd expect from a similar A4 1200-based system. Z3740 seems to perform on par with the A4 5000 (equal or better @ 7-ZIP & PCMark7, worse at CB11.5), so Z3740 gives you 3x better MT CPU performance than the A4 1200. Thank god OEMs probably figured it out.

 

[Abwx]

Worse battery life + total crap CPU performance is what I'd expect from a similar A4 1200-based system. Z3740 seems to perform on par with the A4 5000 (equal or better @ 7-ZIP & PCMark7, worse at CB11.5), we're talking about a 3x MT CPU performance gap. Thank god OEMs probably figured it out.

And what was the power comsumption for thoses perfs
to occur.??.

Strange that Anand made a ton of benchs but did
elude power comsumption tests wich would have showed
much bigger numbers than what was displayed at IDF ,
guess it would have been annoying to display such
oddities, hence the complete silence on the much
hyped power comsumption numbers.

 

[drikkie]

Neverless , the demonstration still stand overhelmingly.

You never take any numbers for granted and now you believe a random marketing quote of 9 hours.... Never have seen any battery numbers that where really usefull in practice.

And with that, screen is ofcourse a little bigger, but total CPU power and probably even GPU is way over the A4 1200.

I don't see any real victory there.

 

[Vesku]

If MSI's 9 hours on 34.78Wh battery is a web browsing workload then the A4-1200 is actually a pretty good 10 inch tablet chip considering AMD is using TSMC 28nm vs Baytrail-T on Intel 22nm.

 

[Abwx]

You never take any numbers for granted and now you believe a random marketing quote of 9 hours.... Never have seen any battery numbers that where really usefull in practice.

And with that, screen is ofcourse a little bigger, but total CPU power and probably even GPU is way over the A4 1200.

I don't see any real victory there.

No victory but nothing as ugly as claimed by a few
members here who are messing wishfull thoughts based
on personal interests with unfounded and unrelentlessly
spread internet urban legends willfully repeated for
deffamations purposes , see the post above about
an alleged 3 hours battery life of AMD s Temash...

Now on the subject of this topic i would imagine
the brouhaha if ever the tables were turned
in respect of said unpublished and still unmeasured
spec that was still hyped as the great thing that
would come with BT.

 

[Abwx]

If MSI's 9 hours on 34.78Wh battery is a web browsing workload then the A4-1200 is actually a pretty good 10 inch tablet chip considering AMD is using TSMC 28nm vs Baytrail-T on Intel 22nm.

Actualy it may do worse in browsing but better
in video playback battery life wise but one has
to consider the 11.6" screen likely consuming more
than the T100 10"1.

 

[frozentundra123456]

If browsing , wich will hardly use more than a core lightly ,
get 8.68h battery life we can assume without much risk
to be significantly off that video playing will hardly last
more than 5 hours or so , i m even sure that Anand has
the numbers but didnt publish them for some reason , heck ,
despite the wonderfull numbers doctored by Intel he didnt
seem to give as much care for what was announced
as the great advantage of this new chip and even relegated
this spzecification to negligible quantity benchmarks wises..

There s no smoke without fire , that s for sure , and all
evidences point to average battery life , far , very far
from what was ultra hyped.

Yea for sure everything that is favorable to Intel is either cheating or some sort of conspiracy. I only gave a more complete quote of the entire paragraph, of which you only picked the most negative part toward the device. Seems ironic that you highlight a phrase taken out of context but refuse to accept a more complete quote because it doesn't fit your agenda.

 

[Abwx]

Yea for sure everything that is favorable to Intel is either cheating or some sort of conspiracy. I only gave a more complete quote of the entire paragraph, of which you only picked the most negative part toward the device. Seems ironic that you highlight a phrase taken out of context but refuse to accept a more complete quote because it doesn't fit your agenda.

So getting power comsumption numbers is cheating because
it could be unfavourable to Intel ?..

Then if theses numbers were so great why didnt Anand
focus on what was presented as the main advantage
of this chip.?..

He told us that at IDF he saw 2.5W for the SoC
when running four cores (at what was seemed to be
2.2-2.3GHz) with CB but still when testing CB perfs
he didnt even bother to check if the numbers did agree.


Actualy what favour Intel is the complete silence
about their chip real TDPs on tablets and this is on
purpose , i never saw any firm keeping silent about
their products specs when they had actual superiority.

The only thging we see are benches where the chip
run far above its ridiculous SDP numbering schemes.

 

[ComplexEntity]

So getting power comsumption numbers is cheating because
it could be unfavourable to Intel ?..

Then if theses numbers were so great why didnt Anand
focus on what was presented as the main advantage
of this chip.?..

He told us that at IDF he saw 2.5W for the SoC
when running four cores (at what was seemed to be
2.2-2.3GHz) with CB but still when testing CB perfs
he didnt even bother to check if the numbers did agree.


Actualy what favour Intel is the complete silence
about their chip real TDPs on tablets and this is on
purpose , i never saw any firm keeping silent about
their products specs when they had actual superiority.

The only thging we see are benches where the chip
run far above its ridiculous SDP numbering schemes.

I hate to bust your BayTrail-5hr-playback-time bubble because apparently some foreign sites had already done their sharing of test on T100 as well. Here is one link in Chinese:

http://pad.pconline.com.cn/359/3596882_all.html#content_page_1

under the battery test section, the author specified "after browsing web, video compressing, photo processing for a total period of 40 min, the tablet dropped 8% of battery" (which is quite close to Anand's 8.68 hrs number), "after playing back a 2.35 GB, 1024x576 mkv video file for 140 min, the battery dropped 31%" (which translates to 140/0.31/60 = about 7.5 hrs playback time). "in our stand by test, when I left work the battery stands at 23%. The next morning when I arrived at work the battery had about 18% left. It is about 5% in 13 hrs". All of the battery related tests were done under "Balanced" mode. Of course, at the last the author also mentioned "under different stress tests battery life may vary".

I don't know about you, but these numbers are pretty damn impressive for a 10.1 in tablet with a 30 whr battery. You can wait for Ananad's final review regards to battery life, but I doubt it will make much of a difference anyway.

 

[Sweepr]

9 hours number is pure MSI marketing (''when reading ebooks or performing word processing'') , you cant compare this crap to retail Bay Trail-T PC numbers. ASUS also says T100 lasts up to 11 hours on a single charge, yet real life numbers are a bit lower (more like ~9 hours web browsing). Anyway, only fanboys would choose a chip with 1/3 of Z3740's MT CPU performance, thats why traditional PC OEMs chose Bay Trail-T for their x86 Windows 8.1 tablets.

Even A4 1200 GPU numbers fail to impress. From MSI's website:

www.msi.com/product/windpad/W20-3M.html

I hate to bust your BayTrail-5hr-playback-time bubble because apparently some foreign sites had already done their sharing of test on T100 as well. Here is one link in Chinese:

http://pad.pconline.com.cn/359/3596882_all.html#content_page_1

under the battery test section, the author specified "after browsing web, video compressing, photo processing for a total period of 40 min, the tablet dropped 8% of battery" (which is quite close to Anand's 8.68 hrs number), "after playing back a 2.35 GB, 1024x576 mkv video file for 140 min, the battery dropped 31%" (which translates to 140/0.31/60 = about 7.5 hrs playback time).

Thanks for the numbers. According to MSI marketing W20's battery life is up to 6 hours when playing 1080p videos.

 

[monstercameron]

9 hours number is pure MSI marketing (''when reading ebooks or performing word processing'') , you cant compare this crap to retail Bay Trail-T numbers. ASUS also says T100 lasts up to 11 hours on a single charge, yet real life numbers are a bit lower (more like ~9 hours web browsing). Anyway, only fanboys would choose a chip with 1/3 of Z3740's MT CPU performance, thats why traditional PC OEMs chose Bay Trail-T for their x86 Windows 8.1 tablets.

Even A4 1200 GPU numbers fail to impress. From MSI's website:

www.msi.com/product/windpad/W20-3M.html

http://tweakers.net/productreview/83223/medion-akoya-e1318t.html

 

[Abwx]

I hate to bust your BayTrail-5hr-playback-time bubble because apparently some foreign sites had already done their sharing of test on T100 as well. Here is one link in Chinese:

http://pad.pconline.com.cn/359/3596882_all.html#content_page_1

under the battery test section, the author specified "after browsing web, video compressing, photo processing for a total period of 40 min, the tablet dropped 8% of battery" (which is quite close to Anand's 8.68 hrs number), "after playing back a 2.35 GB, 1024x576 mkv video file for 140 min, the battery dropped 31%" (which translates to 140/0.31/60 = about 7.5 hrs playback time). "in our stand by test, when I left work the battery stands at 23%. The next morning when I arrived at work the battery had about 18% left. It is about 5% in 13 hrs". All of the battery related tests were done under "Balanced" mode. Of course, at the last the author also mentioned "under different stress tests battery life may vary".

I don't know about you, but these numbers are pretty damn impressive for a 10.1 in tablet with a 30 whr battery. You can wait for Ananad's final review regards to battery life, but I doubt it will make much of a difference anyway.

Would one really serious in measurement protocols
use the device battery state displaying.??..

Any estimation based on this will be extremely flawed ,
remember that the battery discharge is not linear , it could
well , due to manufacturing variations , that the settings
would be very slightly different from a battery to another ,
with some displaying 30% even if it s 20% actualy.

Any serious reviewer would use actual measurement
apparatus and not such quasi irrational and indeed
extremely amateurish sloppy estimation when doing
a power comsumption evaluation.

Edit : Welcome at AT CPU n OCKING...

 

[Abwx]

Thanks for the numbers. According to MSI marketing W20's battery life is up to 6 hours when playing 1080p videos.

And how much for the T100 with Z3740
despite its smaller screen..???..:sneaky:

 

[Abwx]

"after playing back a 2.35 GB, 1024x576 mkv video file for 140 min, the battery dropped 31%" (which translates to 140/0.31/60 = about 7.5 hrs playback time).

An exemple of estimation based on false assumptions...

The first 31% last 140mn BUT you wont get 280mn
when the battery will be at 62% discharge because
the battery internal resistance increase with discharge
so to get your 280mn you ll have to exhaust more like
69% of the battery and it will be even worse when you
get closer to full discharge.

Actualy using only 31% of the battery to make an estimation
is to use the part of the battery life cycle that is the most
favourable to display purely marketing numbers.

Edit : From thoses numbers i speculate that video PB
will last barely 6 hours for said devices once Anand
clear the dust , unless he use a film whose action
occur at nights only...