Intel Broadwell Thread

[coercitiv]

How come idle battery runtime is so important ?? I though Laptops switch to sleep state when idled after a few minutes.

Idle is very important for people doing light work (browsing, writing etc). Look at those Dell Latitude 5550 reviews, the Haswell unit has more stamina under a non-realistic full load scenario, but as you transition to browsing/office work the Broadwell system lasts longer.

 

[AtenRa]

Idle is very important for people doing light work (browsing, writing etc). Look at those Dell Latitude 5550 reviews, the Haswell unit has more stamina under a non-realistic full load scenario, but as you transition to browsing/office work the Broadwell system lasts longer.

Idle is not browsing, writing etc.

Idle (without WLAN, min brightness)

Edit: look again the wifi surfing, you only gain 15min

Battery runtime

Dell Latitude E5550 Core i5 5300U Broadwell
Idle (without WLAN, min brightness) = 12h 01min
WiFi Surfing = 6h 21min
Big Buck Bunny H.264 1080p = 6h 27min
Load (maximum brightness) = 1h 06min


Dell Latitude E5550 Core i5 4310U Haswell
Idle (without WLAN, min brightness) = 8h 43min
WiFi Surfing = 6h 05min
Big Buck Bunny H.264 1080p = 6h 22min
Load (maximum brightness) = 1h 18min

 

[coercitiv]

Idle is not browsing, writing etc.

Of course it is, while we read and write CPUs idle. (pun intended )

Edit: look again the wifi surfing, you only gain 15min

It's only 15 minutes because it's a scripted scenario, I bet the browser loads pages and scrolls them faster than a human would. But more important than that, it has gone from -12 minutes under full load to +15 minutes while browsing.

 

[Abwx]

The laptop with the Core M has 10% lower perf/Watt, since idling times are more than compensated by very few loading times the browsing test is a good average of thoses two states, there s no differencs and this was already visible in Intel slides for whom has paid attention.

 

[Fjodor2001]

It's only 15 minutes because it's a scripted scenario, I bet the browser loads pages and scrolls them faster than a human would.

These tests are usually designed to mimic average user browsing behavior. Here are the Test criteria descriptions from the Notebookcheck review site:

http://www.notebookcheck.net/Our-Test-Criteria.15394.0.html

On WiFi browsing mode it says:

WiFi mode: the possible battery life while surfing the Internet via WiFi with medium brightness (~150 cd/m²) and power-saving options ("Power Saver" mode) switched on. We measure the run time by letting the device run an automatic script, which picks a mix of websites and switches between them every 40 seconds.

So I would say that's quite normal average browsing behavior for a human...

 

[tential]

How come idle battery runtime is so important ?? I though Laptops switch to sleep state when idled after a few minutes.

Because you aren't 100% of the time utilizing the PC? Do you guys just simply pretend to not understand how people use computers when you make your arguments?

Lets say I'm taking notes during a meeting. I may not be typing/web browsing the whole time. Maybe only once every 1 minute or so. So the PC sits idle for a good while each time I'm not interacting with it. 33% increase in idle time means that that portion of the time I'm spending idle is now much less of a drain on my battery than before.

Of course Idle battery time is therefore important and no one should have to explain this to you. This should be fairly obvious.

 

[coercitiv]

So I would say that's quite normal average browsing behavior for a human...

My premise is the script constantly scrolls the page, while a human would rather do it in a non linear fashion, stopping along the way. Moreover, browsing also means reading, and sometimes people spend even minutes reading the content of a screen / page, not just 40 seconds. Other usage patterns might be considerably more aggressive though, there's no script to cover them all, and in this regard notebookcheck's take is as good as any.

 

[Fjodor2001]

My premise is the script constantly scrolls the page, while a human would rather do it in a non linear fashion, stopping along the way.

Nothing in the Test criteria indicates this as far as I can see.

Moreover, browsing also means reading, and sometimes people spend even minutes reading the content of a screen / page, not just 40 seconds.

On the other hand it is also quite normal to click around a few links and just briefly access web pages for a few seconds on each page. On the average, spending 40 seconds on a page seems quite normal to me.

Sure, depending on exact browsing behavior the difference in battery life might be both shorter and longer. But I think the test Notebookcheck does is good enough as an indication of expected average web browsing behavior over WiFi. So the point AtenRa made is valid, which is the essence of all of this.

 

[NTMBK]

I would bet very serious money that Broadwell-U 2+2 and Broadwell-Y are identical dice.

Sure, but back then BDW-U wasn't announced, so they didn't disclose anything about that. Since it's the same 2+2 die, one would assume that Intel didn't bother creating 2 version of the same silicon. Why would they do that? Maybe that's the reason BDW-U doesn't clock higher in boost.

Hmm, good point that BDW-U wasn't announced at that point! In that case I agree with you guys, and it probably is the same process for -U and -Y. I guess the full-fat high performance variant is reserved for BDW-K and SKL-S? (And the Xeons?) I also suspect that the Xeon D will use the higher density/lower power variant, given that it is going after the microserver market.

 

[coercitiv]

I think the test Notebookcheck does is good enough as an indication of expected average web browsing behavior over WiFi. So the point AtenRa made is valid, which is the essence of all of this.

It is valid, no sense in me pursuing this any further. I was debating variance around a data point, not the data itself.

 

[oobydoobydoo]

Idle is not browsing, writing etc.



Edit: look again the wifi surfing, you only gain 15min

Battery runtime

Dell Latitude E5550 Core i5 5300U Broadwell
Idle (without WLAN, min brightness) = 12h 01min
WiFi Surfing = 6h 21min
Big Buck Bunny H.264 1080p = 6h 27min
Load (maximum brightness) = 1h 06min


Dell Latitude E5550 Core i5 4310U Haswell
Idle (without WLAN, min brightness) = 8h 43min
WiFi Surfing = 6h 05min
Big Buck Bunny H.264 1080p = 6h 22min
Load (maximum brightness) = 1h 18min

So I am really confused by this. Idle power time increases by a very respectable 37.8% and load increases by 18.1%, which is good, but wifi browsing decreases? Maybe there is some issue with this model laptop

Otherwise, the increases seem pretty good with battery life. Performance may not have increased much at all, but if we remove the outlying wifi results the increase in battery life is pretty good (18-37.8%).

Edit:
The increase is only 37.8%, not 40%. 721min/523min = 1.378585. The other figure comes from 78min/66min = 1.181818

Still very confused.

 

[AtenRa]

So I am really confused by this. Idle power time increases by a very respectable 37.8% and load increases by 18.1%, which is good, but wifi browsing decreases? Maybe there is some issue with this model laptop.

Otherwise, the increases seem pretty good with battery life. Performance may not have increased much at all, but if we remove the outlying wifi results the increase in battery life is pretty good (18-37.8%).

Edit:
The increase is only 37.8%, not 40%. 721min/523min = 1.378585. The other figure comes from 78min/66min = 1.181818

Actually load battery runtime decreases with Broadwell comparing those two Laptops.

 

[Galatian]

Actually load battery runtime decreases with Broadwell comparing those two Laptops.

Which does makes sense...kinda. Race to idle and all. Broadwell clocks much higher with Boost.

 

[coercitiv]

Which does makes sense...kinda. Race to idle and all. Broadwell clocks much higher with Boost.

Actually they have similar turbo boost clocks, but Broadwell honors its max clocks all the time, while Haswell chip downclocks under extreme stress.

This may also be related to how Dell configured the unit, afaik Intel allows OEMs to customise this behaviour to some extent.

 

[Galatian]

Actually they have similar turbo boost clocks, but Broadwell honors its max clocks all the time, while Haswell chip downclocks under extreme stress.

This may also be related to how Dell configured the unit, afaik Intel allows OEMs to customise this behaviour to some extent.

You are right: 2,9 vs 3,0 GHz Turbo. But that would mean that Broadwell doesn't get as hot with the same TDP, so it can Turbo longer?

 

[Idontcare]

Actually load battery runtime decreases with Broadwell comparing those two Laptops.

Don't worry, at some point these folks will slowly, but surely, start to realize that there is a darned good reason why Intel has delayed and delayed their full rollout of 14nm broadwell SKUs...and in the meantime have gone silent on 10nm.

Meanwhile the data we do have are rather telling, as you rightly point out.

 

[Abwx]

You are right: 2,9 vs 3,0 GHz Turbo. But that would mean that Broadwell doesn't get as hot with the same TDP, so it can Turbo longer?

With the same TDP it has lower perfs, the two exemples above HW is consuming up to 24W while BDW is close to 29W, that is at SoC level.

And both are supposed to be 15W chips...

 

[coercitiv]

You are right: 2,9 vs 3,0 GHz Turbo. But that would mean that Broadwell doesn't get as hot with the same TDP, so it can Turbo longer?

No, at least not with the data from this review. Under full load the Broadwell equipped Latitude consumes more power than the Haswell unit and runs a bit hotter/louder too. We could argue it is total system power and other components might be at fault, although they are quite similar and idle power data even tilts in favor of the Broadwell system.

Moreover, judging from the reviews, the Haswell CPU goes as high as ~70C while Broadwell is alowed to go up to ~80C. As I said in my previous post, it may also be related to how Dell configured the unit.

My take on all of this is Broadwell U is more efficient than Haswell U, but the difference may not be as significant as we were led to believe.

 

[coercitiv]

HW is consuming up to 24W while BDW is close to 29W, that is at SoC level.

Where are you getting this from?

 

[Abwx]

Where are you getting this from?

From thoses reviews, also notice how the fan is spinning hard at full load, 46.4dB to exhaust the 43.5W total comsumption, HW is at 37W total power drain..

http://forums.anandtech.com/showpost.php?p=37137214&postcount=1677

 

[coercitiv]

From thoses reviews, also notice how the fan is spinning hard at full load, 46.4dB to exhaust the 43.5W total comsumption, HW is at 37W total power drain..

Fine, but how did you end up with HW @ 24W and BDW @ 29W, both at SoC level?

Please don't tell me you substracted idle system power draw from full load system power draw.

 

[oobydoobydoo]

Actually load battery runtime decreases with Broadwell comparing those two Laptops.

You're right! For some reason I just assumed the longer load time was for Broadwell.


I have to agree with Idontcare; This is very telling. Now the outlier appears to be idle power usage and it otherwise appears that Broadwell actually somehow decreased efficiency from Haswell. This is not looking good. I will edit my mistake in my last post.

 

[Abwx]

Fine, but how did you end up with HW @ 24W and BDW @ 29W, both at SoC level?

Please don't tell me you substracted idle system power draw from full load system power draw.

Not exactly, but that s close once you factored the delta by 0.81 and substracted 2-2.5W, and that s assuming that the SoC consume nothing at idle....
Where do you think that all thoses watts are going..?.

For comparison here a laptop with an actual 15W TDP APU :

http://www.notebookcheck.net/HP-Pavilion-13-a093na-x360-Convertible-Review-Update.130928.0.html

23.8W at full load and 19.8W delta, do the maths..

 

[coercitiv]

Not exactly, but that s close once you factored the delta by 0.81 and substracted 2-2.5W, and that s assuming that the SoC consume nothing at idle....
Where do you think that all thoses watts are going..?

Screen - 4W delta from min to full brightness on my ~300cd 15" IPS
RAM - 3W max delta from idle to load reported via HWMonitor on 2x8GB DDR3@1600
Fan - 2W delta from powered off to full throttle

So that's 9W I could think off excluding any increase in power draw from the MB itself. A bit more than 2-2.5W

 

[Abwx]

Screen - 4W delta from min to full brightness on my ~300cd 15" IPS
RAM - 3W max delta from idle to load reported via HWMonitor on 2x8GB DDR3@1600
Fan - 2W delta from powered off to full throttle

So that's 9W I could think off excluding any increase in power draw from the MB itself. A bit more than 2-2.5W

If you want to go in the details to grab a few watts i can point that there s only 8MB on the laptop as well as AC/DC and DC/DC conversion being better than 90% efficency at power like 40W, and so on and we ll get back to twice the chip official TDP, the 29W figure is close to reality, perhaps 27-28W but not less.