Intel Broadwell Thread

Page 58 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Arkaign

Lifer
Oct 27, 2006
20,736
1,377
126
Unless I was running a literal 7" tablet, I'd pick the i5-5200U 10,000 times out of 10,000 over A8X. The whole reason to run a laptop, even an ultralight like the Air, is to have adequate power to multitask in a full fledged OS with full fledged apps.

Now that doesn't mean that Apple couldn't put together some kind of jumbo A-series CPU and get something more of that caliber ready to go. But I find the idea that Apple would somehow trash their legacy OSX compatibility in pursuit of a slower tablet processor in the MBA sort of ludicrous. Whether we like it or not, OSX and the apps out there are entrenched in x86, and benefit greatly from the performance of big chips and big chips cut down a bit to fit the midrange TDP of a 11"-13" frame. With that form factor, dropping to 4W or whatever would make for insane battery life, but also drag the hell out of running iTunes while running Safari with a bunch of tabs, while dabbling with some other apps. As it is now, the Air is borderline with performance, dropping CPU grunt significantly would upset many many people.

This will be something to revisit in a few years though, it certainly isn't going to happen out of the blue anytime soon.
 
Mar 10, 2006
11,715
2,012
126
Edit: before I leave for work, is anybody impressed with this CPU? I get that you guys think Apple will keep Intel, fair enough you may be right, but what do you personally think of the performance increases Intel is giving today? Especially in graphics?

I think it's very telling that Intel is keeping Skylake on schedule. Broadwell is good chip in that it's better than Haswell, but as has been discussed in this thread, Intel is missing a number of key features. I'm also pretty disappointed that Intel hasn't made the U and Y family of chips full SoCs...having an on-package 32nm PCH can't be good for power efficiency.

I think at the very least, Skylake's PCH will be 22nm which should bring lower power consumption.
 

Sweepr

Diamond Member
May 12, 2006
5,148
1,142
131
Apple has already matched Haswell-U performance with A8X at much better power draw. That is all they need for a retina air.

Just curious, what are you basing this claim on and what exactly are you talking about (CPU/GPU)?
Core i5 5200U was ~10-20% faster than Core i5 4200U @ 3DMark Physics and CB11.5, that's more than the IB -> Haswell transition for 15W SKUs and quite impressive for a Tick+.





Let's hope Apple pulls out something new and interesting like the Lenovo Lavie Z, packing Broadwell-U (not Core M) and a 13'' 2560x1440 IGZO screen (optional) in a 1.72 pound device. Their current Macbook Air looks like a dinosaur with its 1440x900 TN screen and ~3 lb (edit).
 
Last edited:

Eug

Lifer
Mar 11, 2000
23,752
1,285
126
You are absolutely spot on with the lack of H.265 in Gen. 8. This is a pretty serious feature omission that everybody from Apple to MediaTek has in next generation SoCs. Intel remedies this with Skylake, though.

At this point, I wouldn't be surprised to see Apple just swap in Broadwell-U into the current MBA chassis and save the big redesign for a Skylake MBA launched in time for the back to school shopping season.
Yeah, my iPad already supports HEVC (with FaceTime), and my next iPhone (probably a 6s) will support it as well.

As for my MacBook, I'd like to see a new form factor with Broadwell U if possible, if only because it will likely be intentionally crippled in another important way, and they would likely fix that in the Skylake release.

Let's hope Apple pulls out something new and interesting like the Lenovo Lavie Z, packing Broadwell-U (not Core M) and a 2560x1440 IGZO screen (optional) in a 1.72 pound device. Their current Macbook Air looks like a dinosaur with its 1366x768 TN screen and ~3 lb.
The 3 lb MacBook Air is 13" 1440x900. The 1366x768 version is 11.6" and 2.4 lb.

Mind you I dislike both of them. I hate the screen on the 11.6" since it has poor viewing angles (as bad as the old white MacBooks), and the screen on the 13" has an odd resolution for OS X, and other than resolution, has an image quality that basically rivals my TN 13" MacBook Pro and nothing more.
 

Essence_of_War

Platinum Member
Feb 21, 2013
2,650
4
81
Their current Macbook Air looks like a dinosaur with its 1366x768 TN screen and ~3 lb.

In fairness to the MBA, the 11'' one has the 1366x768 panel, and it weighs 2.4 lbs, the 13'' is 3lbs but has the 1440x900 panel.
 

rootheday3

Member
Sep 5, 2013
44
0
66
Maybe Ice storm is getting close to being CPU bound and it's time to move to benchmarks like Cloud Gate or Sky Diver?
 

Piroko

Senior member
Jan 10, 2013
905
79
91
I hope that the list price of those i3s is an error, there's a huge gap between 160$ and 275$ and they are terrible value at those list prices.
 

mikk

Diamond Member
May 15, 2012
4,173
2,211
136
So more than half what the 50,000 the 28W5500-U scores in cloudgate! How much power do you you think it used?


i7-5500U is a 15W SoC not 28W!


[/IMG]

Shouldn't the HD5500 be demolishing the HD4400??

The HD5500 has more EUs and a more modern design too.


I wrote that Icestorm Unlimited is useless for a GPU comparison and you are asking me this question now. CPU Performance is a big factor. For example 4770k+HD4600 is only 25% slower than A10-7850K according to GPU Scores of Icestorm Unlimited. Intels slide is dumb because they are claiming up to % by using this benchmark for gaming purposes. Intels launch is a mess. Dumb slides and no real world benchmarks, not even a meaningful preview.
 

TreVader

Platinum Member
Oct 28, 2013
2,057
2
0
i7-5500U is a 15W SoC not 28W!





I wrote that Icestorm Unlimited is useless for a GPU comparison and you are asking me this question now. CPU Performance is a big factor. For example 4770k+HD4600 is only 25% slower than A10-7850K according to GPU Scores of Icestorm Unlimited. Intels slide is dumb because they are claiming up to % by using this benchmark for gaming purposes. Intels launch is a mess. Dumb slides and no real world benchmarks, not even a meaningful preview.

I didn't catch that.



My point is not that A8X somehow competes with head-to-head with intel offerings!


My point is that by building A8X, Apple have demonstrated that they are capable of building chips that perform as good or better than intel on a performance per watt basis. Not only that, but they can scale that performance up to 1.5Ghz, which IMO demonstrates a Macbook Air type device is feasible.


That's all I'm saying about Apple's abilities at this point. However, I do think intel has demonstrated very little improvement in performance. They are repeating the bulldozer debacle, except this time it will be called "THE X86 DEBACLE".
 

USER8000

Golden Member
Jun 23, 2012
1,542
780
136
I wrote that Icestorm Unlimited is useless for a GPU comparison and you are asking me this question now. CPU Performance is a big factor. For example 4770k+HD4600 is only 25% slower than A10-7850K according to GPU Scores of Icestorm Unlimited. Intels slide is dumb because they are claiming up to % by using this benchmark for gaming purposes. Intels launch is a mess. Dumb slides and no real world benchmarks, not even a meaningful preview.

The CPU in the Core i5 5200U not only has IPC improvements over the one in the i5 4200U, but has a much higher base clockspeed and is made on a newer node too. So even with a more powerful CPU,it appears the improvement in the GPU section are minor??
 
Last edited:

Essence_of_War

Platinum Member
Feb 21, 2013
2,650
4
81
Not only that, but they can scale that performance up to 1.5Ghz, which IMO demonstrates a Macbook Air type device is feasible.



How are you concluding this? Are you just looking at the base clock of the MBA, the base clock of the ipad, and figuring ~1.5 GHz therefore they can be swapped?
 

witeken

Diamond Member
Dec 25, 2013
3,899
193
106
Maybe Ice storm is getting close to being CPU bound and it's time to move to benchmarks like Cloud Gate or Sky Diver?

IIRC you are an Intel GPU architect or software guy, so I guess your claim is backed up by some information.
 

III-V

Senior member
Oct 12, 2014
678
1
41


How are you concluding this? Are you just looking at the base clock of the MBA, the base clock of the ipad, and figuring ~1.5 GHz therefore they can be swapped?
TreVader consistently has shown he cannot be rational when it comes to Apple and Intel... why he is still allowed to post in the CPU forum, I will never know.
 

TreVader

Platinum Member
Oct 28, 2013
2,057
2
0


How are you concluding this? Are you just looking at the base clock of the MBA, the base clock of the ipad, and figuring ~1.5 GHz therefore they can be swapped?

What do you mean "swapped"? Apple doesn't just "swap" one thing for another and call it a day. They aren't Samsung.


They would have to build a whole new device around the Apple processor. That in no way means it is not possible. If you are trying to argue that the thermal envelope wouldn't allow you to clock it that high, you should check the TDPs. Macbook Airs would allow you to double the cores and up the clock speed on A8X. Look at the difference in TDP between a 3.2Ghz quad core and a 1.6Ghz dual core, it's almost exactly double.
 

mikk

Diamond Member
May 15, 2012
4,173
2,211
136
The CPU in the Core i5 5200U not only has IPC improvements over the one in the i5 4200U, but has a much higher base clockspeed and is made on a newer node too. So even with a more powerful CPU,it appears the improvement in the GPU section are minor??


You really don't understand how it works. Let's say you have a 10% faster CPU and 30% faster GPU and are using a bench which is bottlenecked by both CPU and GPU. Overall maybe 15% faster in this benchmark (depending on what is the bigger bottleneck). So yes, even with a more powerful CPU the benchmark difference is or can be minor. But this not an isolated GPU benchmark and won't give you a representative picture about GPU improvements. As I said Icestorm Unlimited is useless. Furthermore some scores are odd. 17W Ivy Bridge ULV only 5% slower....
 

USER8000

Golden Member
Jun 23, 2012
1,542
780
136
You really don't understand how it works. Let's say you have a 10% faster CPU and 30% faster GPU and are using a bench which is bottlenecked by both CPU and GPU. Overall maybe 15% faster in this benchmark (depending on what is the bigger bottleneck). So yes, even with a more powerful CPU the benchmark difference is or can be minor. But this not an isolated GPU benchmark and won't give you a representative picture about GPU improvements. As I said Icestorm Unlimited is useless. Furthermore some scores are odd. 17W Ivy Bridge ULV only 5% slower....

Broadwell is using a 14NM process node meaning it should have much less issue boosting to higher clockspeeds than the previous generation chips. The base CPU clockspeeds for the 5200U are 33% higher plus the IPC increase on top of this. Looking at the physics score you can see at least a 20% increase,meaning that the 4200U is still sustaining a lower base clockspeed due to thermal limitations(the IPC increase is minor). The IGP would be boosting higher IMHO especially since it is not massively larger in EU count over the Haswell equivalent.

Plus IB and Haswell were made on the same 22NM process node,and remember Haswell had the voltage regulation on-die which probably adds to the TDP(hence I would expect IB to throttle less in such a situation).

So even with a 25% increase in EU count and improvements to the design,I suspect Intel are hitting bandwidth limitations just like with other companies too. This has been masked by the fact Intel started at a lot lower level performance wise with their IGPs,but now performance has reached a decent enough level,we are starting to see other factors come into play too.

It will be interesting to see how much of an improvement the 48 EU parts with L4 cache will have will have over the 40 EU parts with L4 cache.

Personally I expect the IGP performance improvements for the non-L4 cache parts to be minor overall,but battery life should be improved.

However,that is what Broadwell is about - improving battery life.
 
Last edited:

TreVader

Platinum Member
Oct 28, 2013
2,057
2
0
You really don't understand how it works. Let's say you have a 10% faster CPU and 30% faster GPU and are using a bench which is bottlenecked by both CPU and GPU. Overall maybe 15% faster in this benchmark (depending on what is the bigger bottleneck). So yes, even with a more powerful CPU the benchmark difference is or can be minor. But this not an isolated GPU benchmark and won't give you a representative picture about GPU improvements. As I said Icestorm Unlimited is useless. Furthermore some scores are odd. 17W Ivy Bridge ULV only 5% slower....

tl:dr "It looks slow, but it's actually super fast. Trust us."
 

oobydoobydoo

Senior member
Nov 14, 2014
261
0
0
I didn't catch that.



My point is not that A8X somehow competes with head-to-head with intel offerings!


My point is that by building A8X, Apple have demonstrated that they are capable of building chips that perform as good or better than intel on a performance per watt basis. Not only that, but they can scale that performance up to 1.5Ghz, which IMO demonstrates a Macbook Air type device is feasible.


That's all I'm saying about Apple's abilities at this point. However, I do think intel has demonstrated very little improvement in performance. They are repeating the bulldozer debacle, except this time it will be called "THE X86 DEBACLE".

A8X is a good processor, but there is no way they can replace an x86 cpu with an ARM one that easily. You are wrong here. Intel still has the best process in the business, they just haven't yet adjusted the architecture!. Remember how Ivy Bridge was only 5%, but haswell was up to 15%? Changes in architecture now offer much more in performance than process improvements, which mostly offer density improvements. You will see a big improvement with skylake.


I agree that intel has slowed down, but come on! Let's give the team that made sandy bridge a chance (they are making skylake!).
 

Roland00Address

Platinum Member
Dec 17, 2008
2,196
260
126
So finally core m devices that actually make sense.

$699 T300 Chi, 1080p ips, 4gb memory, 128gb ssd, keyboard dock. Using the cheaper core m sku but supposedly is faster than the yoga.

So windows with a faster cpu in a form factor similar to the ipad air (it is heavier and wider bug similar)

The T300 is also cheaper than the haswell i3 surface pro 3, is faster (though closer to the i3 than the i5), comes with the keyboard, 128 storage vs 64, and is a $100 cheaper. The surface pro by contrast has the digitzer pen but loses the included keyboard in its price. Oh supposedly the t300 will have an lte option.

There is nothing wrong with core m besides the price of the devices and hopefully the t300 is a trend and not the cheapest. Sure it is no i5 haswell u, but remember it is an awesome chip and broadwell u will be 50 to 60% faster but Broadwell U has 3 times the tdp at 15w vs 4.5w.
 

Dufus

Senior member
Sep 20, 2010
675
119
101
Why is that i7-5600U only scoring 3.1 when the lower clocked i7-5500U scores 3.3? Hopefully it's not another example of Lenovo throttling performance.

When are manufacturers going to take the TDP as a specification and not as the maximum amount of power that processor can use. Plenty of thermal headroom from what I have seen so allow 20-25W for the 15W SKU and while on power saver limit to 15W. IMO would be a lot less users disappointed with performance as were with Gen 4.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |