Intel Broadwell Thread

[North01]

Good point, but I don't see much evidence of significant score differences between 3.2 and 3.3.

Here is a Venue 7140 with 5y71 on 3.1.5

Single-Core: 2539
Multi-Core: 4406

http://browser.primatelabs.com/geekbench3/1739600

And here is it with the 5y10 on 3.2.2

Single-Core: 1939
Multi-Core: 4098
http://browser.primatelabs.com/geekbench3/1621607

The point is that there doesn't appear to be a trend of score deflation going from 3.2 to 3.3. Thus, there isn't much reason to think that the Asus score is inflated because it was running on the older version.

Sorry if I created any confusion, my main point was to show how ASUS' implementation is likely better than Dell's. The additional reference was just to show the performance difference between the 32-bit and 64-bit tests, not to compare version numbers. I just happened to share the version number to put as much information out there (for those that might be interested).

 

[mikk]

http://www.notebookcheck.com/Test-Acer-Aspire-V3-371-58DJ-Notebook.134784.0.html

Here is another BDW ULT test. HD Graphics 5500 is roughly 20% faster. I hope Skylake improvements are bigger than this.

 

[witeken]

You'd think that the -30% power of 14nm, the Gen8 architecture and the 4 extra EUs would result in a higher 15W efficiency than 1.2x Haswell...

 

[mikk]

It's pretty clear that Gen9 is Intels real next gen GPU. Furthermore I think edram is required to make a big difference. Memory bottlenecks getting bigger and bigger even on those ULTs.

 

[witeken]

Remember this article?

http://www.extremetech.com/computin...ape-indicates-major-improvements-over-haswell

In general terms, Intel’s Ben Widawsky, who works on Intel’s Linux graphics driver efforts, says that “Broadwell graphics bring some of the biggest changes we’ve seen on the execution and memory management side of the GPU… [the changes] dwarf any other silicon iteration during my tenure, and certainly can compete with the likes of the gen3->gen4 changes.”

 

[Arachnotronic]

It's pretty clear that Gen9 is Intels real next gen GPU. Furthermore I think edram is required to make a big difference. Memory bottlenecks getting bigger and bigger even on those ULTs.

Yep, and this is confirmed by all of the leaked slides/CPU World reports that Skylake-ULT will feature 64MB of eDRAM.

 

[rootheday3]

Remember this article?

http://www.extremetech.com/computin...ape-indicates-major-improvements-over-haswell

I think the comments about the extent/scope of Gen8 overhaul relate to things like:
1) per-process graphics page tables
2) shared virtual memory (including but not limited to OCL2.0)
2a) support for IA32 page tables (see next item)
2b) coherent memory between graphics and cpu (no flushes needed)
3) workload submission model

 

[witeken]

I think the comments about the extent/scope of Gen8 overhaul relate to things like:
1) per-process graphics page tables
2) shared virtual memory (including but not limited to OCL2.0)
2a) support for IA32 page tables (see next item)
2b) coherent memory between graphics and cpu (no flushes needed)
3) workload submission model

What does that mean in more layman terms?

What's the scope of Gen9?

 

[Enigmoid]

http://www.notebookcheck.com/Test-Acer-Aspire-V3-371-58DJ-Notebook.134784.0.html

Here is another BDW ULT test. HD Graphics 5500 is roughly 20% faster. I hope Skylake improvements are bigger than this.

Interesting 18W package TDP.

2.5 ghz prime 95 + 900 mhz igp.

31.5W total power under prime + furmark

 

[Fjodor2001]

Strange. I thought Broadwell was going to be the BIG improvement for Intel when it comes to iGPU performance.

Now it's Skylake instead? What happened to providing major iGPU uArch changes in ticks, and CPU uArch changes in tocks?

 

[III-V]

Strange. I thought Broadwell was going to be the BIG improvement for Intel when it comes to iGPU performance.

Now it's Skylake instead? What happened to providing major iGPU uArch changes in ticks, and CPU uArch changes in tocks?

Nobody really knows, because AMD/Intel/Nvidia/etc. have stopped communicating things.

 

[Fjodor2001]

I knew they are no longer telling much in advance any longer. But now they are not even telling once the products have been released, as for Broadwell?

 

[ShintaiDK]

Broadwell gave a huge improvement in sampler performance.

You will notice that Broadwell IGP performance is much more consistent than Haswell. And it was a bottleneck to be fixed before moving on.

 

[Fjodor2001]

But would you consider the Broadwell iGPU performance improvements in general to be THE big improvement for Intel CPUs? I.e. that it stands out among other improvements we've seen lately going from one generation to the next?

If anything, I would instead say the opposite - that we got a rather small iGPU performance improvement in Broadwell...

 

[NTMBK]

But would you consider the Broadwell iGPU performance improvements in general to be THE big improvement for Intel CPUs? I.e. that it stands out among other improvements we've seen lately going from one generation to the next?

If anything, I would instead say the opposite - that we got a rather small iGPU performance improvement in Broadwell...

To be fair, Intel GPUs have gotten past the low-hanging fruit of "clearly broken" performance issues- expect more of these focused performance improvements as time goes on, instead of a "big bang".

 

[Fjodor2001]

To be fair, Intel GPUs have gotten past the low-hanging fruit of "clearly broken" performance issues- expect more of these focused performance improvements as time goes on, instead of a "big bang".

In that case, does it mean that they will never catch up with AMD in IGP performance? Because if they'd like to do that, they have to keep increasing performance at a faster pace than AMD.

 

[Enigmoid]

As far as the igp goes people are always comparing intel's to AMD's without looking at the elephant in the room : Drivers.

Intel doesn't release performance drivers for games (or much) yet games are what people are always comparing intel's igp with. Imagine if AMD or Nvidia never did any code optimization for games and games never got patched to run on AMD/Nvidia hardware.

For instance, take a game like DA:I, FC4, Crysis 3, BF4 and play it on pre-launch AMD or Nvidia drivers. Now play it on current drivers. There is a huge amount of performance gains and other bug fixes that are addressed with later drivers. Intel doesn't do that to any large extent (you can see this occasionally where all of a sudden intel will drop in performance compared to AMD in some game or gain a ton of performance and suddenly become nearly competitive).

Therefore if you want to look at something like architectural efficiency intel's igp needs a handicap.

 

[Fjodor2001]

http://appleinsider.com/articles/15...air-refresh-with-intel-broadwell-chips-feb-24

Apple may be planning to debut a relatively minor update for its existing MacBook Air lineup in a matter of days, though won't be the anticipated 12-inch redesign with a high-resolution Retina display, according to a new rumor.
[...]
The alleged MacBook Air refresh will feature improved Intel processors for a speed bump, according to Letemsvetemapplem.eu Specifically, Apple could choose the 2.2GHz Core i5-5250U, a 1.8GHz Core i5 chip, or a 2GHz Core i7 variety.

Not quite like Apple to release a product that will be obsolete within 3 months though (when Skylake is supposed to be released)...

 

[Sweepr]

Perhaps they are holding back the major Mabook Air update till Skylake-U. I wouldn't blindly trust that VR-Zone roadmap, I still expect Skylake-U to be released in late Q3 or Q4.

Even if the new Air doesnt pack a Retina display I hope they update those old 1440x900 TN panels. I'd take a slim and very well built Dell XPS 13 (2015) with a gorgeous 3200x1800 IPS panel (2000:1 contrast) (10 hours battery life @ WiFi) over a current Macbook or a minor Broadwell updated version any day.

 

[Arachnotronic]

Perhaps they are holding back the major Mabook Air update till Skylake-U. I wouldn't blindly trust that VR-Zone roadmap, I still expect Skylake-U to be released in late Q3 or Q4.

Even if the new Air doesnt pack a Retina display I hope they update those old 1440x900 TN panels. I'd take a slim and very well built Dell XPS 13 (2015) with a gorgeous 3200x1800 IPS panel (2000:1 contrast) (10 hours battery life @ WiFi) over a current Macbook or a minor Broadwell updated version any day.

I'm actually a bit surprised that Apple is able to keep gaining significant share with the current Air models despite the ancient display. I think people buy Macs for the Mac OS experience + industrial design + brand + customer support, not necessarily for the technical bells as whistles.

Oh, and those amazing trackpads

 

[Sweepr]

I'm actually a bit surprised that Apple is able to keep gaining significant share with the current Air models despite the ancient display. I think people buy Macs for the Mac OS experience + industrial design + brand + customer support, not necessarily for the technical bells as whistles.

Oh, and those amazing trackpads

Yes, brand loyalty aside they do have nice build quality and a functional trackpad. After using a 1080p IPS notebook since 2013 (ASUS UX32VD) I am very exigent on display quality. Fast forward to 2015 and you can buy a 1080p IPS display Chromebook for $329 while QHD+ displays are starting to come down to reasonable prices (1440x900 TN is a no go for a +$999 notebook at least for me).

 

[jdubs03]

Yes, brand loyalty aside they do have nice build quality and a functional trackpad. After using a 1080p IPS notebook since 2013 (ASUS UX32VD) I am very exigent on display quality. Fast forward to 2015 and you can buy a 1080p IPS display Chromebook for $329 while QHD+ displays are starting to come down to reasonable prices (1440x900 TN is a no go for a +$999 notebook at least for me).

Yea a 1440x900 was nice with my 2006 edition Macbook Pro . 9 years later, there is no way that is acceptable.

 

[Sweepr]

Pipo’s new Core M tablet costs ~$399 (Windows & Android dual-boot)

Pressure from Google and Microsoft might stop big-name manufacturers like Asus from building dual-boot devices. Smaller Chinese OEMs, like Pipo, couldn’t care less.

Pipo just introduced a new Atom-powered Windows 2-in-1 last month, and now they’re adding another new device to their line-up. This time it’s a 10.1-inch tablet powered by a 14nm Intel Core M processor, and it does, in fact, dual-boot both Windows 8.1 and Android. It’s expected to retail for ¥2500, which is about $399.

Apart from its 10.1-inch display and the M-5Y10C chip, Pipo’s latest also features a generous 4GB of DDR3 RAM. There’s a single USB 3.0 port and SD card slot for expansion, a dock connector for a click-in keyboard. A separate power plug means one more hole in the casing, but also slightly speedier charging.

http://liliputing.com/2015/02/pipo-core-m-tablet-dual-boots-windows-and-android.html
https://translate.google.com/transl...alla-de-101%E2%80%B3-y-4gb-de-ram/&edit-text=

That's an incredible price for a Core M tablet. It's also smaller than most Core M devices out there, I was waiting for a 10'' tablet instead of the usual 12-13'' convertibles. I have some Windows licences here so (to me) it doesnt really matter if its activated or not.

 

[witeken]

$400? That's a good $200 below Intel's number and even more above people's expectations.

 

[Sweepr]

$400? That's a good $200 below Intel's number and even more above people's expectations.

Indeed, that's a lot of bang for the buck. If Pipo can do it it's just a matter of time till we see more $399-599 Core M devices from other companies.