Intel Broadwell Thread

[witeken]

Skylake-K does not represent the full Skylake line-up. Everything Intel said about Skylake is that it will be released in 2015. That doesn't mean every Skylake part will be released and BK also explained that it typically takes 1.5 years. It was Cannonlake that isn't changed.

 

[Sweepr]

Asus unveils Zenbook UX305 Broadwell-powered ultrabook

Asus is updating its Zenbook line of premium thin and light laptops with a new model featuring Intel’s new low-power Core M “Broadwell” processor.

The Asus Zenbook UX305 is a 13.3 inch laptop that measures less than half an inch thick and weighs just over 2.6 pounds.

The notebook has a 13.3 inch, 3200 x 1800 pixel display and supports up to 256GB of solid state storage. Asus says the notebook should get over 10 hours of battery life. It features HDMI output, 802.11ac WiFi, 3 USB ports, a headset jack, and an SD card reader.

Mobile Geeks got a chance to check out the new Zenbook and reports it has a matte display, Bang & Olufsen speakers, and comes in black or white colors.

http://liliputing.com/2014/09/asus-unveils-zenbook-ux305-broadwell-powered-ultrabook.html

Impressive ultrabook from ASUS, one of the first Broadwell-based PCs announced. Can't wait for M$'s future ''Broadwell'' Surface Pro.

 

[SAAA]

Umm... Nope. It is only recently that we found out Intel would release the gimped S version and remain silent on the status of Skylake K for 2015. If you look at the roadmap, Skylake K should be a Summer/Fall 2015 product.

I7 860 launched September 2009.
New architecture: 2600K launched January 2011.
New architecture: 4770K launched June 2013.

Intel vowed to release a new architecture every 2 years or so. That means if Skylake-K launches in 2016, it's 1 year late. And actually, it was always meant to be that Skylake K would replace Broadwell-K in 2015 which aligns with the 2 year cadence. Intel said the delay of Broadwell won't impact Skylake which is now total BS because if what they said is true, we would have seen Skylake-K June 2015-December 2015 at the latest!

Broadwell's delays clearly impacted Skylake K or otherwise Intel wouldn't be launching Broadwell K and holding back Skylake K in 2015.
---

I wouldn't mind if Intel pushed the enthusiast X99 and subsequent replacement lines to enthusiasts. For me a $340 quad-core seems like a big rip-off now that Intel is selling an underclocked hexa-core for barely more. Right now X99 platform isn't as attractive as can be due to slightly higher DDR4 prices. However, if Intel can bring out the latest architecture 6-8 months after the mainstream platform in the form of $390 6-core, the mainstream i7 is dead to me because sooner or later DDR4 will reach pricing parity with DDR3.

I think Intel will move the mainstream i7 below $300 in a generation or 2 OR they will have to keep these 500mhz+ clock delta between the i5 and i7s. Once DDR4 reaches pricing parity, it will cost just $120-140 more to go full hexa-core over the Z107 platform.

Agree with the above. Also I understand that each architecture and family of processors cost money/time to develop but really is it going to change anything if Intel releases all Skylakes for desktops and Broadwell only for mobile? Heck they are even missing the discussed FIVR and they will probably be much better overall, making more people upgrade for once.

Releasing Broadwell-K instead seems like a ripoff for enthusiasts, that's understandable for the hex+ line, derived from server versions, but why that should influence common desktop parts?
Plus the IGP affair that is so interesting right now, will it ever be useful other than for graphics... certainly not in Broadwell because that's mostly a shrunk Haswell and you are paying even more die for nothing. What is it now, 50% with GT3 variant?

Maybe it's all a plan to voluntarily make the quad core line uninteresting so all the attention moves to hex cores and pricer Broadwell parts... those will probably be also less expensive for them without that huge IGP on die.

 

[witeken]

SAAA, you will be able to buy an LGA Skylake SKU for desktop in (presumably) Q2 of next year, so that makes your whole comment pretty much irrelevant.

 

[mikk]

I don't understand the fuss. We know that Devil's Canyon's successor will be released in Q2. So unlocked Skylake will be released in 2016, no surprises here.

Exactly this, there is nothing new. We know how the Roadmap looks. Furthermore the Roadmaps is not necessarily an indicator of CPU Performance, the price is important asd well. Because of the GT3 SKU with edram and OC support BDW-K should be considerably pricier and in the Roadmap above Skylake-S, there really is no surprise here.

 

[liahos1]

Asus unveils Zenbook UX305 Broadwell-powered ultrabook



http://liliputing.com/2014/09/asus-unveils-zenbook-ux305-broadwell-powered-ultrabook.html

Impressive ultrabook from ASUS, one of the first Broadwell-based PCs announced. Can't wait for M$'s future ''Broadwell'' Surface Pro.

can that really be classified as an ultrabook. i thought it didnt have a touch screen

 

[RussianSensation]

SAAA, you will be able to buy an LGA Skylake SKU for desktop in (presumably) Q2 of next year, so that makes your whole comment pretty much irrelevant.

No, it's not irrelevant. You keep defending Intel and won't admit that Intel is now facing a 1 year slip with both Broadwell-K and Skylake-K. Since this is an enthusiast site where a lot of us overclock, we really don't care about gimped 65W Skylake-S.

I just priced out 5820K vs. 4790K systems and it looks like even before DDR4 prices drop, the mainstream Intel i7 is already in trouble.

i7 4790K = $297 after tax
Asrock Z97 Extreme 6 = $123 after tax and rebates
2x4GB DDR3 = $85
Total: $505

i7 5820K = $318 after tax
Asrock X99 Extreme 4 = $212 after taxes
2x4GB DDR4-2400 = $117
Total: $647

Difference of only $142 for 2 more cores which is way more future proof over the next 4-5 years since not only do you get a system that's more well-rounded for work/multi-threaded tasks and gaming, but you can also run dual-SLI and M.2 drive with 32Gb/sec bandwidth, something you can't do an 99% of Z97 boards.

Broadwell-K i7 at $340 is dead in the water next fall unless it overclocks to 5.0Ghz on a $30 cooler and/or Intel delays BW-E 6820K. Intel needs Skylake-K for the mainstream i7 K series to make sense moving forward.

 

[witeken]

The way I read his comment was that he thought Skylake-S wouldn't be released for desktops, which is not true. I never said that Broadwell-K and it successors are not 1 year too late, nor am I defending it. This post pretty much sums up my opinion.

 

[liahos1]

Some Core M Devices being announced at IFA besides the ASUS Sweepr mentioned

http://www.anandtech.com/show/8483/ifa-2014-lenovo-launches-more-pcs-from-laptops-to-tabletops

"Lenovo announced an updated version of the Helix. The new model is 12% lighter than the outgoing version, coming it at 1.8 lbs, and at 0.38 inches thick, it is 15% thinner than the first gen device. The processors are getting a big boost, with Intel Core M CPUs now under the hood which should boost performance, but even more so, battery life and thermals"


http://liliputing.com/2014/09/hp-envy-x2-2-1-tablets-broadwell-chips-outed-bit-early.html

"It looks like HP may be preparing to launch a new line of 2-in-1 tablets that become notebooks with the aid of tablet docks. German website WinFuture spotted a few retail listings for unannounced products called the HP Envy 13 x2 and HP Envy 15 x2.

Both computers seem to be Windows machines with full HD displays and Intel Core 5y10 M Broadwell processors."

surprisingly they are all 5y10 or 5y10a skus

https://www.youtube.com/watch?v=AfPruPJlcew - Chippy says the base clock is 900mhz, so a 100mhz bump from the cpu world article

http://www.cpu-world.com/news_2014/2014070201_Details_of_first_Broadwell_Y_mobile_processors.html

maybe the the turbo's clock higher than the cpu-world article as well.

Looking forward to a 5y70 based 2-1.

 

[jdubs03]

maybe the the turbo's clock higher than the cpu-world article as well.

Looking forward to a 5y70 based 2-1.

This for sure. 5Y10's aren't going to be that powerful, so I hope we see more of the 70's, especially for something like the Helix, which is a device I have always liked (despite the price).

The issue with all of these new devices is I can't see them holding up the same performance levels as their previous U-series counterparts. We might be talking about a performance decrease of 25% compared to U-series.

 

[liahos1]

This for sure. 5Y10's aren't going to be that powerful, so I hope we see more of the 70's, especially for something like the Helix, which is a device I have always liked (despite the price).

The issue with all of these new devices is I can't see them holding up the same performance levels as their previous U-series counterparts. We might be talking about a performance decrease of 25% compared to U-series.

saw this on twitter

Chippy ‏@chippy 43m
Messing with a Core M I got, from a single core at 50% load, 843 on Sunspider. 1.0Ghz Clock. Scope for a lot more there.

anybody understand these results? I am not really familiar with sunspider but dont the baytrail devices score significantly lower (better)?

http://www.anandtech.com/show/7314/intel-baytrail-preview-intel-atom-z3770-tested/3

as does the i3 surface?

http://www.anandtech.com/show/8287/first-look-the-799-microsoft-surface-pro-3-with-core-i3

 

[jdubs03]

saw this on twitter

Chippy ‏@chippy 43m
Messing with a Core M I got, from a single core at 50% load, 843 on Sunspider. 1.0Ghz Clock. Scope for a lot more there.

anybody understand these results? I am not really familiar with sunspider but dont the baytrail devices score significantly lower (better)?

http://www.anandtech.com/show/7314/intel-baytrail-preview-intel-atom-z3770-tested/3

as does the i3 surface?

http://www.anandtech.com/show/8287/first-look-the-799-microsoft-surface-pro-3-with-core-i3

Sunspider is a very short test so I wouldn't put too much emphasis on that, but despite that 843ms is slow, I'm sure the Y-series Haswell were faster than that. Core i5-4202Y was 204.5ms on average according to http://www.notebookcheck.net/Intel-Core-i5-4202Y-Notebook-Processor.102728.0.html. So yeah maybe this a bad bench, I'm not sure, but that is very disappointing if this is a common result.

 

[Fjodor2001]

I fear this before but next year is looking like a yawnfest for the mainstream Intel desktop CPU platform. I didn't quite understand why Intel would launch Broadwell-K to only later introduce the superior Skylake-K. Well, it looks like my fears are coming true and Intel will gimp Skylake in 2015 and launch the slower Skylake-S, while Broadwell-K is what will replace i7 4790K. If so, I don't foresee any worthwhile Skylake-K in 2015 at all. Perhaps, Intel is shifting the performance aspect to the X99 platform with Broadwell-E taking over HW-E in 2015 as the performance offering.

http://www.fudzilla.com/home/item/35645-broadwell-to-be-faster-than-skylake-s-in-desktop

I guess Intel wants to recoup the R&D costs on Broadwell which more or less means Skylake-K is likely going to be pushed into 2016. I am now just counting until an article pops up stating just that.

So Intel will intentionally be releasing Skylake with lower performance than possible. Nice. Not. :thumbsdown:

I guess it's to avoid the embarrassment of skipping desktop Broadwell completely due to the delays now putting it in Skylake timeframe.

 

[Fjodor2001]

"Intel’s 14nm Core M detailed: Broadwell delivers huge performance, dramatic power reduction"

http://www.extremetech.com/computin...creases-dramatically-better-power-consumption

Not sure about the "huge performance" though. They're talking about 2-3% CPU performance increase over Haswell, and 20-30% for the iGPU.

I think where Broadwell will shine instead is in low power (~4.5 W) chips. I.e. low power for the performance it provides in that TDP range. Focusing on that likely made Broadwell non-optimal for high performance desktop use.

 

[Homeles]

So Intel will intentionally be releasing Skylake with lower performance than possible. Nice. Not. :thumbsdown:

I guess it's to avoid the embarrassment of skipping desktop Broadwell completely due to the delays now putting it in Skylake timeframe.

If they give Broadwell and Skylake-K the same treatment as Devil's Canyon... it's a non-issue. My concern with DC is that Intel has to keep up with it, and give that special treatment from then on out... otherwise, they're going to go backwards at some point.

"Intel’s 14nm Core M detailed: Broadwell delivers huge performance, dramatic power reduction"

http://www.extremetech.com/computin...creases-dramatically-better-power-consumption

Not sure about the "huge performance" though. They're talking about 2-3% CPU performance increase over Haswell, and 20-30% for the iGPU.

I think where Broadwell will shine instead is in low power (~4.5 W) chips. I.e. low power for the performance it provides in that TDP range. Focusing on that likely made Broadwell non-optimal for high performance desktop use.

At the same TDP, "huge performance" is a given. At least, if you consider ~30% to be a "huge performance" gain, as it has become these days.

 

[Vesku]

I thought Broadwell's jump in iGPU was going to be more than +20% when it came to EU count. Guessing Gen 8 EUs may be a decent bit larger on die than Gen 7 for even more performance per EU than expected, after all normally a 20% increas in just EU count wouldn't translate directly into a ~20-30% performance boost. Would mean the individual EU going from Gen 7 to Gen 8 has improved between 30-50% which sounds to me like quite a bit more transistors involved.

saw this on twitter

Chippy ‏@chippy 43m
Messing with a Core M I got, from a single core at 50% load, 843 on Sunspider. 1.0Ghz Clock. Scope for a lot more there.

anybody understand these results? I am not really familiar with sunspider but dont the baytrail devices score significantly lower (better)?

http://www.anandtech.com/show/7314/intel-baytrail-preview-intel-atom-z3770-tested/3

as does the i3 surface?

http://www.anandtech.com/show/8287/first-look-the-799-microsoft-surface-pro-3-with-core-i3

Is he saying a Broadwell 1GHz core working at 50% (some sort of power setting cap or software implemented cap?) scores 843? If so then it's at least faster than Ivybridge, my IB system at 3.8GHz scores 142.8ms if I linearly extrapolate it would be ~1085 at 1GHz and 50% load. So 'on paper' it seems Haswell->Broadwell IPC improvement will be similar to IvyBridge->Haswell.

 

[jdubs03]

I thought Broadwell's jump in iGPU was going to be more than +20% when it came to EU count. Guessing Gen 8 EUs may be a decent bit larger on die than Gen 7 for even more performance per EU than expected, after all normally a 20% increas in just EU count wouldn't translate directly into a ~20-30% performance boost. Would mean the individual EU going from Gen 7 to Gen 8 has improved between 30-50% which sounds to me like quite a bit more transistors involved.



Is he saying a Broadwell 1GHz core working at 50% (some sort of power setting cap or software implemented cap?) scores 843? If so then it's at least faster than Ivybridge, my IB system at 3.8GHz scores 142.8ms if I linearly extrapolate it would be ~1085 at 1GHz and 50% load. So 'on paper' it seems Haswell->Broadwell IPC improvement will be similar to IvyBridge->Haswell.

I think maybe they are referring to IPC improvements, because their slide with using less than 50% power yet performing 20% better just doesn't seem possible, especially if they're allocating more space to the GPU. If we take that 50% load at 1.0Ghz, we get 421.5ms there, and let's assume its a 5Y10 with 2.0Ghz turbo, that gets down to 210.75ms, which is quite good for such a low TDP. However, there are too many variables still in flux to determine anything at all performance wise at this point, IDF may give us a peek at a reference device with performance numbers, or so I hope.

 

[JDG1980]

According to Chiploco, three Broadwell-Y CPUs have already been discontinued by Intel... prior to their actual release. This includes the 5Y10 in the Asus products announced recently. I can't help but wonder if something is going very wrong with Intel's new process... no one else can do 20nm and below with any kind of reliability, maybe Intel can't either?

 

[Homeles]

According to Chiploco, three Broadwell-Y CPUs have already been discontinued by Intel... prior to their actual release. This includes the 5Y10 in the Asus products announced recently. I can't help but wonder if something is going very wrong with Intel's new process... no one else can do 20nm and below with any kind of reliability, maybe Intel can't either?

It seems that everything is revolving around that stepping.

 

[pw257008]

Seems TSX-related, not 14nm-related

 

[jdubs03]

It seems that everything is revolving around that stepping.

That was my first thought as well. Anything after December should be good with Core M.

 

[liahos1]

http://www.pcworld.com/article/2602840/dell-packs-core-m-chip-in-latitude-13-7000-tabletlaptop.html

Another core m device

 

[jpiniero]

That seems really weird though; so the launch products won't have TSX but shortly thereafter they will get a new stepping that will have the fix.

 

[Abwx]

http://www.hardware.fr/news/13878/core-m-fin-vie-avant-leur-lancement.html

.

 

[Vesku]

That seems really weird though; so the launch products won't have TSX but shortly thereafter they will get a new stepping that will have the fix.

They thought they had TSX covered until a developer discovered an issue with the implementation that apparently can not be fixed purely through microcode. Even Intel can't magically fix already manufactured products.

"Intel finds specialized TSX enterprise bug on Haswell, Broadwell CPUs"
http://www.pcworld.com/article/2464...enterprise-bug-on-haswell-broadwell-cpus.html