Intel Broadwell Thread

[witeken]

Who cares about TDP when it seems iPair Air battery life is as good or better than the only existing Core M machine available? The TDP number in isolation is a completely useless number in particular in "closed" devices like this.

Anyway let's wait for other Core M devices before spitting at the CPU face, IMHO Yoga 3 is just a bad design

The TDP number does not include power consumption of a 13.3" 6MP (probably non-IGZO) panel.

FYI, the Yoga Pro 3 has 88% more screen area and if it doesn't use IGZO, the power consumption per area will be 2x more than the iPad.

 

[Nothingness]

The TDP number does not include power consumption of a 13.3" 6MP (probably non-IGZO) panel.

FYI, the Yoga Pro 3 has 88% more screen area and if it doesn't use IGZO, the power consumption per area will be 2x more than the iPad.

And in the end I don't care: if they didn't use a proper screen technology to reduce power consumption and increase battery life, then their device is not worth my money. And puts a bad picture on Core M by the way...

All I say is that in the end if the device as a whole stinks, it stinks, no matter what TDP CPU x or y has

 

[Qwertilot]

While we're being reasonable, a rather expensive laptop really should have a much bigger battery in it than an ipad
(Hugely more space for it.).

 

[Roland00Address]

Witiken and Nothingness you are both right. Having a much higher amount of pixels and a 67% larger screen sucks for battery life and screen consumption. (Witiken's point). But Nothingness is also right at the end of the day I do not care why the product sucks, if it it sucks it just sucks. Lenovo drop the ball on the engineering on this computer, they should have made a much more balanced design for a 1k laptop. The engineer inside of me wants to throw things at Lenovo for messing this up.

Lenovo management though does not care, they went with pretty specs and pretty watch hinges for they just want to sell the laptop to you. Appearance and large specs numbers sell computers sadly. They do not want to make a balanced device that you will love and adore months later. They just want you to take home the product and once you take it home you are very unlikely to return it.

Are OEMs really surprised that people do not trust them to make high end products anymore and thus their brands image is trash? If people are going to pay 1k+ for a computer they go with macs for macs usually do not have such glaring flaws but people have been burned enough times with lenovos, dells, hps, etc which have glaring flaws in a 1k+ laptop.

 

[Arachnotronic]

While we're being reasonable, a rather expensive laptop really should have a much bigger battery in it than an ipad
(Hugely more space for it.).

The laptop industry in general is a joke. There's always a compromise that needs to be made on the Windows PC side; always something not quite right about a given system, even at the high end.

That is why I now use a Retina MacBook Pro. It has a great screen, great touchpad, fast processor, great keyboard, and an attractive industrial design.

 

[Fjodor2001]

I'd really like to see an Apple based product based on Broadwell-M. They rarely screw up like OEM Windows PC laptop manufacturers. So an Apple based product is more likely to show off what Broadwell-M really can achieve.

What worries me though is that Apple has not announced any 12" fanless MBA based on Broadwell-M as has been expected. Could it be that even Apple has given up on producing a fanless MBA/Ultrabook based on Broadwell-M? Oh well, maybe I'm worrying too much...

 

[Nothingness]

I'd really like to see an Apple based product based on Broadwell-M. They rarely screw up like OEM Windows PC laptop manufacturers. So an Apple based product is more likely to show off what Broadwell-M really can achieve.

What worries me though is that Apple has not announced any 12" fanless MBA based on Broadwell-M as has been expected. Could it be that even Apple has given up on producing a fanless MBA/Ultrabook based on Broadwell-M? Oh well, maybe I'm worrying too much...

Two hypotheses come to mind:

1. Intel yield issues with 14nm meant they could not have delivered enough chips to Apple.
2. it's not only Yoga 3 that stinks, and Core M isn't good enough for even MBA.

 

[inf64]

Has this been posted on this forums already?
http://pclab.pl/art60181.html

It's Intel Core M 5Y70 review. Doesn't look that impressive versus older gen of Core series but dominates Baytrail(as expected).

 

[Nothingness]

Has this been posted on this forums already?
http://pclab.pl/art60181.html

It's Intel Core M 5Y70 review. Doesn't look that impressive versus older gen of Core series but dominates Baytrail(as expected).

To sum up (and provided I correctly understood): the 5Y70 (1.1GHz, turbo up to 2.6GHz, TDP 4.5W) lies performance-wise between 4210Y (1.5GHz, turbo up to 1.9GHz, TDP 11W, SDP 6W) and 4020Y (1.5GHz, no turbo, TDP 11W, SDP 6W). Basically no performance improvement, it looks like Core M can't use its turbo for long in the tested machine (again a Yoga 3 it seems).

Now let's wait and see how battery performs on models using that CPU

 

[frozentundra123456]

It depends on actual power usage. Come on, you are comparing a 4.5w TDP chip vs 11w and getting similar performance. That is a huge increase in performance per watt, if the respective TDPs are comparable.

Even if you compare Broadwell TDP to Haswell U SDP, there is a considerable improvement in performance per watt.

 

[Enigmoid]

What worries me though is that Apple has not announced any 12" fanless MBA based on Broadwell-M as has been expected. Could it be that even Apple has given up on producing a fanless MBA/Ultrabook based on Broadwell-M? Oh well, maybe I'm worrying too much...

Apple rarely announces anything before they start selling it. Not too much of a big deal IMO.

 

[Roland00Address]

Has this been posted on this forums already?
http://pclab.pl/art60181.html

It's Intel Core M 5Y70 review. Doesn't look that impressive versus older gen of Core series but dominates Baytrail(as expected).

Another yoga 13 review. Remember on its numbers Core M can have one of three tdp targets.

3.5 Watts
4.5 Watts (29% higher tdp allowed)
or 6 Watts (71% higher tdp allowed)

Apparently Lenovo with the yoga 3 pro chose the lowest of the tdp options the 3.5w option (why? I really can't guess though more on that later). So core m with a 3.5w tdp is getting performance between the 6w core i3 4020y (1.5 ghz) and the 6w core i5 4210y (1.5ghz with turbo 1.7ghz with both cores and 1.9 ghz with one core) is still a big improvement since you cut the tdp from 6w to 3.5w and enable better form factors for now you can do razor thin tablets.

I also want to see the gpu performance for these 3.5w and 6w tdp parts. Newer gpu architecture that is tweaked for low tdps.

Now personally I can care less about razor thin tablets so I want to see what core m can do at 6 watts. I understand why people would want the lower tdp tablets but if the device is going to have a keyboard I can care less if the tablet portion is 7mm, 8mm, 11mm, or even higher. Hell I wish intel allowed tdps even higher than 6w for tablets that can be docked or have a fan like the surface pro 3.

Now the lenovo 3 pro is 12.8 mm thick when closed with both slices. I do not know how thick is the keyboard portion which houses the soc. I do know the keyboard portion is thicker than the screen portion. My question is did Lenovo neuter their convertible to get a desired thickness.

Thickness numbers are from notebook check with the exception of the Yoga Pro 3 which is from lenovo

• 17 mm Yoga 13
• 15.6mm Yoga 11 (tegra 3)
• 17.5 mm Yoga 11s (ivybridge 13w)
• 16 mm Yoga Pro 2 13” (haswell 15w)
• 19.3 mm Thinkpad Yoga 12.5" (haswell 15w) the thinkpad model has a better keyboard, retractible keys, and an optional digitizer supposedly it is more durable than the Pro 2 for the business crowd
• 17.5 mm Yoga 11s (haswell 6w)
  vs
• 12.8 mm Yoga pro 3.

So did we get a 3.5w sku since we settled on shaving 3 or 4 mm from the thickness? Picts of showing how the yoga pro 3 thickness is comparing keyboard vs screen Screen halfs on top of each other . Screen halfs side by side.

Intel slide showing what tdps are needed for which tablets sizes and thickness

 

[Nothingness]

It depends on actual power usage. Come on, you are comparing a 4.5w TDP chip vs 11w and getting similar performance. That is a huge increase in performance per watt, if the respective TDPs are comparable.

Even if you compare Broadwell TDP to Haswell U SDP, there is a considerable improvement in performance per watt.

Read my posts above: who cares of the improved efficiency if the CPU is put in an overpriced with poor battery life laptop? You can appreciate how much you want Core M efficiency, right now we have no end-user product that is worth its money.

Core M still doesn't have a product that shows how much it is good. Or perhaps as Intel17 hypothesized no such product will ever exist.

 

[IntelUser2000]

Apparently Lenovo with the yoga 3 pro chose the lowest of the tdp options the 3.5w option

I am starting to agree with Abwx regarding the Core M.

If its truly running at 3.5W, why does it need a fan?

Even if you compare Broadwell TDP to Haswell U SDP,

An Intel presentation I've seen shows that it performs same, but at 4.5W for the Core M instead of 6W for the Core i7 4610Y: http://www.chip.com.my/wp-content/uploads/2014/10/intel-core-m-cpu-malaysia-launch-5.png

Look at this, its the Yoga 3 Pro:

PC applications: http://pclab.pl/art60181-2.html
Mobile/Synthetic applications: http://pclab.pl/art60181-3.html


Sucks at real world mimicking benchmarks, and really good at synthetic(ie. fake) or really bursty mobile benchmarks. I can't believe how much of a difference it is! In PC applications, it loses against even Bay Trail! Yet it turns around and crushes Bay Trail in mobile benchmarks. Benchmarking chip is what it is.

I think when Intel said Core M is a "mobile" chip, it means it can't run anything other than mobile applications. It would have worked for a $20 chip that fits in an 8-inch Tablet.

 

[AtenRa]

All im waititng is for gaming benchmarks, runing Cinebench or x264 etc on such devices is only academic.

It is nice to be fast in browsing but the extra high price for that alone will only make the Core M look bad against 14nm ATOMs and 20nm competition.

 

[SAAA]

I am starting to agree with Abwx regarding the Core M.

If its truly running at 3.5W, why does it need a fan?
http://www.chip.com.my/wp-content/uploads/2014/10/intel-core-m-cpu-malaysia-launch-5.png
http://pclab.pl/art60181-3.html

You can put even a 1W processor in a plastic box and it will overheats very fast if the walls can't dissipate 1J of energy every seconds. Same box made of metal may work fine.

So apparentely that design needs a fan even at 3.5W because it can't dissipate enough passively, crap there isn't even an heatsinkand and you want it fanless?
I built torch at home using a 5W led, it overheated after 1-2 minutes (at 30% efficiency it produces ~3.5W of heat) so I decided to put an aluminium panel behind the led and finally it kept a reasonable temperature, hot but manageable.

 

[AtenRa]

You can put even a 1W processor in a plastic box and it will overheats very fast if the walls can't dissipate 1J of energy every seconds. Same box made of metal may work fine.

So apparentely that design needs a fan even at 3.5W because it can't dissipate enough passively, crap there isn't even an heatsinkand and you want it fanless?
I built torch at home using a 5W led, it overheated after 1-2 minutes (at 30% efficiency it produces ~3.5W of heat) so I decided to put an aluminium panel behind the led and finally it kept a reasonable temperature, hot but manageable.

Heat-sink fan (heat-pipe) is clearly visible in that image (top left corner).

Edit: second image added

 

[AtenRa]

Question, 4.5W TDP is the total for the package including both CPU + PCH ??

 

[witeken]

Question, 4.5W TDP is the total for the package including both CPU + PCH ??

Yes, since Haswell-U.

 

[AtenRa]

Yes, since Haswell-U.

ok thx

 

[zir_blazer]

I don't recall if anyone mentioned this, but...

Xeon E3-1200 v4 series will be available sometime in the middle of 2015, and it will be compatible with current Denlow platform and C220 series chipsets.

C22x Chipsets were released on Haswell launch date alongside the Desktop 8-series lineup (Z87 and so on), and its pretty probable that they're all based on the same silicon. As all LGA 1150 Chipsets but H97 and Z97 were available on the market from months before the announcement that Broadwell would requiere new Chipsets, this news is literally stating that there are high chances that Broadwell could pretty much work on any LGA 1150 Motherboard. Didn't Intel said that Devil's Canyon also needed the new H97 and Z97 Chipsets yet earlier Motherboards got support for it, too?

 

[crashtech]

I don't recall if anyone mentioned this, but...


C22x Chipsets were released on Haswell launch date alongside the Desktop 8-series lineup (Z87 and so on), and its pretty probable that they're all based on the same silicon. As all LGA 1150 Chipsets but H97 and Z97 were available on the market from months before the announcement that Broadwell would requiere new Chipsets, this news is literally stating that there are high chances that Broadwell could pretty much work on any LGA 1150 Motherboard. Didn't Intel said that Devil's Canyon also needed the new H97 and Z97 Chipsets yet earlier Motherboards got support for it, too?

Very speculative, but interesting nonetheless.

 

[Atreidin]

Is there any indication that there will be a Broadwell or Skylake successor to the 4790K? I don't see myself getting any CPUs in the future without VT-d support and it's nice that the highest-performance non-enthusiast processor has it.

 

[inf64]

Is there any indication that there will be a Broadwell or Skylake successor to the 4790K? I don't see myself getting any CPUs in the future without VT-d support and it's nice that the highest-performance non-enthusiast processor has it.

There should be a Broadwell-K part that succeeds 4790K.

 

[IntelUser2000]

So apparentely that design needs a fan even at 3.5W because it can't dissipate enough passively, crap there isn't even an heatsinkand and you want it fanless?

Which manufacturer attaches a fan FIRST before attaching a heatsink?? When have you ever heard of a CPU not needing a fan nevermind a heatsink? That was more than 30 years ago! EVERY CPU in the world needs a heatsink. That's just assumed by default. The difference is whether it requires a FAN or not.

Education: http://stech3.firstpost.com/wp-content/uploads/2014/09/Intel-Core-m-1.jpg

Right from the horse's mouth. Actually, the Yoga 3 Pro design(and every laptop) uses a heatPIPE. That has a super low C/W rating. That means they are lot better than regular heatsinks for cooling.

I built torch at home using a 5W led, it overheated after 1-2 minutes (at 30% efficiency it produces ~3.5W of heat) so I decided to put an aluminium panel behind the led and finally it kept a reasonable temperature, hot but manageable.

You still didn't need a fan though. Core M in the Y3P requires heatpipe AND a fan. That makes it questionable what TDP its really running at.

All im waititng is for gaming benchmarks, runing Cinebench or x264 etc on such devices is only academic.

Gaming would make it worse, not better. When its CPU only the CPU has PLENTY of TDP for itself. The GPU is the biggest reason for power use on these chips nowadays: http://www.anandtech.com/show/5878/...ation-realtime-igpu-clocks-on-ulv-vs-quadcore

"average package power of 16.65W and a maximum package power of 18.75W (exceeding the TDP for short periods), with the CPU drawing an average of 3.9W (6.22W max) and the iGPU drawing 9.14W on average (and 10.89W max)"

If it can't run at above 1.5GHz for the CPU alone, I doubt when gaming it can do half that fluently.