Haswell Battery Life? [video]

ttechf

Senior member
Jun 11, 2012
351
12
81
Hey, I just saw this link and I hope it's true.


- http://windows7forums.com/blogs/mik...3d-transistors-hitting-market-ivy-bridge.html


Intel explaining how Haswell mobile processors will be able to have up to 10 days of battery life. I surely hope so because I don't care what anyone says, 4,5,6,7 hours, it's never true and never enough. On high settings which I like for a laptop, that diminishes quick. So what are we looking at? High settings for a laptop. NOT a netbook or an ultrabook but for a laptop, 24 hours? I can live with that all day. That would be awesome.

Any other thoughts on this because Intel didn't clearly state whether the same battery life will be made available on full-fledged laptops too. Because I don't want an ultrabook or netbook to get that battery life for at least 24 hours.



Let's hear what you think or any other info you have. Thanks. : )
 

vshah

Lifer
Sep 20, 2003
19,003
24
81
10 days will be standby life w/ updates, no way 10 days of actual usage.
 

dma0991

Platinum Member
Mar 17, 2011
2,723
1
0
The average laptop that are sold now will do about 6 hours, 8 hours as advertised. Double that, I'll say it is achievable. If they claimed all day battery life (24 hours), I might call it plausible. They said it is 10 days, I call it BS. Standby time might be believable but 10 days of constant operation, with screen turned on is highly unlikely unless you're carrying a suitcase sized battery with you.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
Its standby power.

6a0120a5580826970c015435d26349970c-800wi


Intel just announced that Haswell (due out in 2013) was designed to reduce connected standby power by 30% compared to Sandy Bridge. The Haswell platform as a whole will deliver a 20x decrease in connected standby power, which Intel believes will enable connected standby battery life to last up to 10 days.

The idea here is to be able to put your notebook to sleep and have it continually fetch cloud updates (email, IMs, tweets) for up to 10 days on a single charge.
 
Last edited:

dma0991

Platinum Member
Mar 17, 2011
2,723
1
0
Its standby power.
The article is a little bit misleading.
However, the one area where Haswell is said to compete in a major way is battery life. According to a report by BBC News and Intel's Kirk Skaugen, the goal of Haswell will be to give you 20x more battery life in standby mode and up to 10 days of battery life in constant operation.
 

ttechf

Senior member
Jun 11, 2012
351
12
81
Thanks for shedding light to this everyone. But if they are getting 10-20 days standby, I think they can give us 12-15 hours of actual computing in a REAL laptop. That is reasonable in my view.
 

Arkadrel

Diamond Member
Oct 19, 2010
3,681
2
0
Overall power usage down 30%? thats what it says right?

That means instead of 4hours of 3dmark, you get like 5,6 hours of it.

Standby time of ~10 days is great, but doesnt mean anything if as soon as you use it to watch movies, it runs out of power within 5hours or so.

In short, standby time isnt nearly as important as how long it can run while eg. playing back a movie, or doing some light work based application.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
Overall power usage down 30%? thats what it says right?

That means instead of 4hours of 3dmark, you get like 5,6 hours of it.

Standby time of ~10 days is great, but doesnt mean anything if as soon as you use it to watch movies, it runs out of power within 5hours or so.

In short, standby time isnt nearly as important as how long it can run while eg. playing back a movie, or doing some light work based application.

No, it says the standby power for the CPu is 30% lower. And the entire platforms lower standby power will bring up to 20x longer standby time.

There is no place it says anything about running power.
 

grkM3

Golden Member
Jul 29, 2011
1,407
0
0
Ill let you guys know how well ivy bridge does in an ultrabook,mine is on its way and It says 9 hours of use on a charge and is only 3lbs.

They wont get 30% over ivys and even in the video they say over second gen 32nm sandys.

real life use were looking at 10-12 hours on a charge with haswell and thats with an ssd drive.Thats pretty amazing since these laptops will be under 3lbs and can do pretty much anything you throw at them.
 

ninaholic37

Golden Member
Apr 13, 2012
1,883
31
91
I never use standby, I always use "shut down". It only takes like 5 seconds to boot up anyway. Is this bad? I've always shut down my 386 DOS/Win3.1 black and white Satellite laptop too, and the battery/laptop still works fine after 20 years.
 

Khato

Golden Member
Jul 15, 2001
1,202
240
106
They wont get 30% over ivys and even in the video they say over second gen 32nm sandys.

Didn't most reviews show practically the same idle power usage between SNB and IVB? I know notebookcheck's review of the quad core models actually showed a ~2 watt increase in idle power consumption for IVB, but that's at least in part due to the higher idle frequency - http://www.notebookcheck.net/Review-Intel-Ivy-Bridge-Quad-Core-Processors.73624.0.html Meanwhile Anandtech's desktop IVB review saw a ~2.6 watt drop compared to SNB... which is well within the margins of idle VID variations between samples.

Regardless, the real number of interest is the 20x reduction in platform 'connected standby power' as that implies every component is finally getting the kind of attention it deserves instead of just being the cheapest option available/the way it's always been done.
 

cantholdanymore

Senior member
Mar 20, 2011
447
0
76
The question I have is from the overall power consumption of a laptop that's being actively used, how much is due to processor and how much due to the screen (say a 15" one)?
 

sm625

Diamond Member
May 6, 2011
8,172
137
106
How many hours playing League of Legends... that is how I measure it. Most notebooks wont last 3. It's not a terribly demanding game either. I suspect Diablo 3 times would be about the same, maybe slightly lower.
 

grkM3

Golden Member
Jul 29, 2011
1,407
0
0
Didn't most reviews show practically the same idle power usage between SNB and IVB? I know notebookcheck's review of the quad core models actually showed a ~2 watt increase in idle power consumption for IVB, but that's at least in part due to the higher idle frequency - http://www.notebookcheck.net/Review-Intel-Ivy-Bridge-Quad-Core-Processors.73624.0.html Meanwhile Anandtech's desktop IVB review saw a ~2.6 watt drop compared to SNB... which is well within the margins of idle VID variations between samples.

Regardless, the real number of interest is the 20x reduction in platform 'connected standby power' as that implies every component is finally getting the kind of attention it deserves instead of just being the cheapest option available/the way it's always been done.

I think what they mean is 30% longer usage using the laptop at full load over 32nm sandy bridge.

and 20x power saving while connected sleep mode
 

IntelEnthusiast

Intel Representative
Feb 10, 2011
582
2
0
On my first laptop with an Intel® Pentium® 4 2.8m processor I would get about 1 hour total battery life out of it. My second computer was based around an Intel Core™ 2 Duo P8400 with this laptop I could get about 3 to 4 hours of battery life (about 1 and 1/2 classes of taking notes). Now with my new Ultrabook™ I am getting over 6 hours (sorry havent run it out yet) with my new Asus Zenbook UX31 and about 5 to 7 days on stand-by. I really do love the performance and battery that I am getting from my new laptop.
 

KompuKare

Golden Member
Jul 28, 2009
1,012
923
136
Regardless, the real number of interest is the 20x reduction in platform 'connected standby power' as that implies every component is finally getting the kind of attention it deserves instead of just being the cheapest option available/the way it's always been done.

I think that's exactly it. This may also mean that Intel will no longer get away with charging $50-$60 for a chipset with diminishing features (as more and more stuff gets integrated) build as an afterthough on a process-node a few generations old. I mean the prices for those 65nm SB and IVB chipsets are fairly high for what they are.

Interestingly enough, the actual CPUs have been zipping power for a while but the mobo and chipsets don't always make it that obvious. I'm sure there are others but the only site I found where they measure the actual draw of the CPU and power circuitry is HT4U.net and while they don't have IVB figures (think maybe they need to modify a mobo), even the power hug Bulldozer 8150 only used 4W idling at the desktop.:

http://ht4u.net/reviews/2011/amd_a8_3870k_llano_apu/index10.php
(js chart not a image so I've only linked to the page)

I also seem to recall an AT review of a Core2 chipset where it showed a fair difference between an older chipset and the newer Intel one at a smaller process node.
 

Ferzerp

Diamond Member
Oct 12, 1999
6,438
107
106
The processor itself is only 20-30% of a laptop's power draw. Even with an infinitely efficient processor, you would only increase battery life by 25-42%.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
I think that's exactly it. This may also mean that Intel will no longer get away with charging $50-$60 for a chipset with diminishing features (as more and more stuff gets integrated) build as an afterthough on a process-node a few generations old. I mean the prices for those 65nm SB and IVB chipsets are fairly high for what they are.

No LGA1155 chipset cost that much. Listprice of Z77 is 48$, 43$ for H77 and so on. But nobody pays listprice. They pay below.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,785
136
Regardless, the real number of interest is the 20x reduction in platform 'connected standby power' as that implies every component is finally getting the kind of attention it deserves instead of just being the cheapest option available/the way it's always been done.

That's exactly it: http://forums.anandtech.com/showthread.php?t=2241480

CPUs are at pretty much at the lowest they can be nowadays. It's other parts of the platform that will get massive improvement.

ninaholic37 said:
I never use standby, I always use "shut down".

"Standby" power in this case means something akin to a smartphone, where you do nothing for a few seconds and basically goes "off". It's a huge change from what they do on PCs now. The conventional sleep mode will become obsolete.

grkM3 said:
Ill let you guys know how well ivy bridge does in an ultrabook,mine is on its way and It says 9 hours of use on a charge and is only 3lbs.

There's minimal power reductions for Ivy Bridge compared to Sandy Bridge. The two biggest ones are that it now supports LPDDR3, and that intermediate frequency steps using few % less power.
 
Last edited:

epidemis

Senior member
Jun 6, 2007
796
0
0
CPUs aren't as low as the can. The thermodynamic limit is 10 million times more efficient.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,785
136
CPUs aren't as low as the can. The thermodynamic limit is 10 million times more efficient.

The main computing portion is. Things like Power Gating and package C7 allows for some real low power. Of course, today's CPUs have more than just the compute portions. But Penryn based Core 2's had 300mW idle power use, which is insigificant considering how much power laptops use while the screen is off!

If anybody has a Sandy Bridge or Ivy Bridge based Ultrabook, go into power management and configure it so it stays idle until the battery runs out WITHOUT making it go to sleep. And do it with the display off. See how long it lasts. From reviews I have seen, it should be in the 3-4W range.
 

Magic Carpet

Diamond Member
Oct 2, 2011
3,477
231
106
Hey, I just saw this link and I hope it's true.


- http://windows7forums.com/blogs/mik...3d-transistors-hitting-market-ivy-bridge.html


Intel explaining how Haswell mobile processors will be able to have up to 10 days of battery life. I surely hope so because I don't care what anyone says, 4,5,6,7 hours, it's never true and never enough. On high settings which I like for a laptop, that diminishes quick. So what are we looking at? High settings for a laptop. NOT a netbook or an ultrabook but for a laptop, 24 hours? I can live with that all day. That would be awesome.

Any other thoughts on this because Intel didn't clearly state whether the same battery life will be made available on full-fledged laptops too. Because I don't want an ultrabook or netbook to get that battery life for at least 24 hours.



Let's hear what you think or any other info you have. Thanks. : )
Well, this is all marketing hype and bold claims, until I see an actual product with at least of 24 hours of constant usage (so they claim), I won't bother.

But I already know today, that just won't happen.
 

epidemis

Senior member
Jun 6, 2007
796
0
0
The main computing portion is. Things like Power Gating and package C7 allows for some real low power. Of course, today's CPUs have more than just the compute portions. But Penryn based Core 2's had 300mW idle power use, which is insigificant considering how much power laptops use while the screen is off!

If anybody has a Sandy Bridge or Ivy Bridge based Ultrabook, go into power management and configure it so it stays idle until the battery runs out WITHOUT making it go to sleep. And do it with the display off. See how long it lasts. From reviews I have seen, it should be in the 3-4W range.

Physics do not agree with you:

http://en.wikipedia.org/wiki/Landauer's_principle