Intel Broadwell Thread

[Sweepr]

http://www.pcper.com/reviews/Processors/IDF-2014-Intel-Shows-Core-M-5Y70-Performance-Numbers

Its glorious

Thanks for posting liahos1. Indeed, very impressive numbers. The Cinebench 11.5 score matches a 15W Core i5 4300U (Haswell-U).

Tegra K1, Apple A8, Snapdragon 805 et al. mobile SoCs just got outclassed. Can't believe they delivered this performance @ 4.5W TDP.

 

[jpiniero]

Probe a misprint. This was tested on llama mountain which is 7.2-7.3mm and fabless

Well, the author did mention that the 4330 had a TDP of 3.6x of this; so 15 W is probally not a misprint. It's probably a cTDP up version of the 5Y70.

 

[North01]

More benchmarks - Link

 

[liahos1]

Its unreal I need one asap. Ms needs to put out an sp4 asap!!!!

 

[Arachnotronic]

Core M looks like a great chip. Must admit though that the most impressive silicon I've seen launched in recent days has been that 18C HSW-EP monster.

 

[liahos1]

Well, the author did mention that the 4330 had a TDP of 3.6x of this; so 15 W is probally not a misprint. It's probably a cTDP up version of the 5Y70.

seems off. Basically no gain in perf/watt vs a Haswell-u if true. Seems illogical. Either way I'm getting one

 

[Eug]

Why the hell didn't they test at 4.5 Watt? Would Intel not let them?

 

[Khato]

Well, the author did mention that the 4330 had a TDP of 3.6x of this; so 15 W is probally not a misprint. It's probably a cTDP up version of the 5Y70.

Had they mentioned any cTDP up for Core M though? I was only aware of the single version that supported a cTDP down to 4W.

Also, if it actually was drawing 15W that tablet would have been more than hot to the touch following that round of benchmarks, no? It is unfortunate that tablet skin temperature wasn't mentioned... but this bodes well to the possibility of other review sites getting some time to spend with Core M during the course of IDF.

 

[Enigmoid]

seems off. Basically no gain in perf/watt vs a Haswell-u if true. Seems illogical. Either way I'm getting one

Yeah, looks wrong because there is no way a tablet that thin (with no fan) can run a 15W processor.

Hugely impressive.

If they can truly reduce power that much then the 14 nm server chips are going to be mind blowing.

On another note, given the small dies and low power usage its disappointing there are no 6C mainstream BW chips planned.

 

[Sweepr]

Had they mentioned any cTDP up for Core M though? I was only aware of the single version that supported a cTDP down to 4W.

Yup, doesn't make sense at all. Also they tested the reference Llama Mountain tablet, which is very thin and fanless. Perhaps the autor confused Core-M with Broadwell-U?

 

[Khato]

On another note, given the small dies and low power usage its disappointing there are no 6C mainstream BW chips planned.

Heh, agreed. I believe that you actually could blame the existence of Core M for that >.>

 

[Abwx]

Yeah, looks wrong because there is no way a tablet that thin (with no fan) can run a 15W processor.

Do you think that it can run at 2.2-2.3GHz with a 4.5W TDP.?

 

[Sweepr]

Quick comparison between Macbook Air 2013 and preliminar Core M scores:

Macbook Air 2013 - 3DMark Ice Storm: ~34.000
www.notebookcheck.net/Review-Apple-MacBook-Air-13-Mid-2013-MD760D-A-Subnotebook.96311.0.html

Apple (and others) can now deliver thin'n'fanless tablets/notebooks/convertibles based on the 4.5W TDP Core-M with similar CPU performance to 15W TDP Haswell-U GT3 and considerably better graphics performance. 12'' Retina Macbook Air, anyone?

 

[erunion]

Do you think that it can run at 2.2-2.3GHz with a 4.5W TDP.?

yes, baytrail can turbo higher than that.

 

[Enigmoid]

Do you think that it can run at 2.2-2.3GHz with a 4.5W TDP.?

Intels 15W ULT CPUs can run the cores (no gpu) at 2.1-2.4 ghz (depending on bin and notebook) on Prime95. (The 3630qm runs prime 95 at 3.2 ghz using according to HW monitor around 40W).

Using a efficiency gain of 2x and much lighter loads yes I could see full CPU only ~2 Ghz clocks under sustained MT on something like CB.

The big problem is that the GPU is very power hungry, perhaps broadwell changes that.

 

[Abwx]

Intels 15W ULT CPUs can run the cores (no gpu) at 2.1-2.4 ghz (depending on bin and notebook) on Prime95. (The 3630qm runs prime 95 at 3.2 ghz using according to HW monitor around 40W).

Using a efficiency gain of 2x and much lighter loads yes I could see full CPU only ~2 Ghz clocks under sustained MT on something like CB.

The big problem is that the GPU is very power hungry, perhaps broadwell changes that.

I dont know for the GPU but numbers published for the SoC display about 30% better power usage than previous generation for the most power hungry chart, there s no reason that if a task takes 1.4W for previous gen and 1.1W for Broadwell then one would use 15W to provide the same perfs than the other one would yield at 4.5W.

http://www.hardware.fr/medias/photos_news/00/45/IMG0045510_1.jpg

 

[Sweepr]

Sneak Peak by François: Core M-5Y70 running 3DMark Ice Storm in a 7.4'' fanless tablet config, >50k overall score and doesn't seem to get hot.

www.youtube.com/watch?v=FIKpypEeo7U&feature=youtu.be

60% faster than Tegra K1, ~3.25x faster than Apple A7.

 

[jdubs03]

Yea Core M is quite the chip, that 3Dmark score is quite good especially for a 4.5W chip, if there was any doubt about the GPU perf compared to TK1, there is no arguing the results we have seen (unless you want to talk about process).

I wonder what the deal is with the Cinebench results from the PcPer and HotH reviews. 2.77pts vs. 2.48pts. There is quite a big difference there. The sunspider difference is significant too, 114ms vs 142ms. Either way those minimum results are still outstanding, but nonetheless I'm curious. Can't wait to see an Anandtech (p)review.

*On 3Dmark graphics, the 5Y70 is 39.6% faster than the TK1.

 

[Sweepr]

I wonder what the deal is with the Cinebench results from the PcPer and HotH reviews. 2.77pts vs. 2.48pts. There is quite a big difference there. The sunspider difference is significant too, 114ms vs 142ms. Either way those minimum results are still outstanding, but nonetheless I'm curious. Can't wait to see an Anandtech (p)review.

*On 3Dmark graphics, the 5Y70 is 39.6% faster than the TK1.

Yup, for some reason HotHardware has slightly lower CPU scores but better GPU scores. They got a 3DMark Ice Storm score of nearly 51k (overall) and (at least it looks like) 56.549 pts @ graphics. That would make it >50% faster than Tegra K1's graphics score.

 

[Abwx]

Yea Core M is quite the chip, that 3Dmark score is quite good especially for a 4.5W chip,

There s an article at AT about the new power management wich is quite different from previous implementations, now a 4.5W TDP chip can perform like a 15W one, no doubt...

UPDATE: A previous version of this graph showed the TDP for the Intel Core M 5Y70 as 15 watts, not the 4.5 watt listed here now. The reasons are complicated. Even though the Intel Ark website lists the TDP of the Core M 5Y70, Intel has publicly stated the processor will make very short "spikes" at 15 watts when in its highest Turbo Boost modes. It comes to a discussion of semantics really. The cooling capability of the tablet is only targeted to 4.5-6.0 watts and those very short 15 watt spikes can be dissipated without the need for extra heatsink surface...because they are so short.

http://www.pcper.com/reviews/Processors/IDF-2014-Intel-Shows-Core-M-5Y70-Performance-Numbers

 

[Eug]

So I don't understand why the 4.5 W is not just SDP then.

 

[Abwx]

So I don't understand why the 4.5 W is not just SDP then.

Because 4.5W is not the power usage in common tasks.

http://www.hardware.fr/medias/photos_news/00/45/IMG0045510_1.jpg

The notion of SDP should be parralleled with a published TDP but now it seems that it s not enough, a third thermal definition is needed....

 

[jdubs03]

Because 4.5W is not the power usage in common tasks.

http://www.hardware.fr/medias/photos_news/00/45/IMG0045510_1.jpg

The notion of SDP should be parralleled with a published TDP but now it seems that it s not enough, a third thermal definition is needed....

Indeed, that 15W is essentially a very short cTDP up. But either way, this has got some serious muscle.

 

[erunion]

http://www.pcper.com/reviews/Processors/IDF-2014-Intel-Shows-Core-M-5Y70-Performance-Numbers

That PCPer explanation is bogus.

TDP is TDP. Turbo boost can exceed TDP for short periods, that true of other Intel chips as well. So its real TDP remains 4.5W.

Indeed, that 15W is essentially a very short cTDP up. But either way, this has got some serious muscle.

Its not cTDP either, which is a configurable TDP. Its just turbo boost working as intended.

 

[ams23]

Sneak Peak by François: Core M-5Y70 running 3DMark Ice Storm in a 7.4'' fanless tablet config, >50k overall score and doesn't seem to get hot.

www.youtube.com/watch?v=FIKpypEeo7U&feature=youtu.be

60% faster than Tegra K1, ~3.25x faster than Apple A7.

Keep in mind that 3DMark Ice Storm Unlimited performance is very CPU-limited. On Tegra K1, the fps in game test 1 is > 200fps and the fps in game test 2 is > 100fps, so this test is not particularly stressful on the latest and greatest ultra mobile GPU's.

The GPU perf. of this Core M variant seems to be within ~ 10% of the Core i5 variant in the Surface Pro 3. Naturally this is very impressive given the lower power consumption. But the GPU perf. difference between Core M and Tegra K1 may not be significant with a more GPU-intensive benchmark such as GFXBench 3.0: