What's so great about Skylake?

/crazymeme.jpg/

Wi-Fi cards consume at MOST 2W at peak load, realistically under a watt. Microwaves run at 1500-2000W. We get more radiation from the sun in an hour than we would from wireless devices in a lifetime. Avert your fears to something more meaningful and realistic like automobile wrecks and cancer due to carcinogens not wireless.

 

Doesn't really seem like there's much to wait for - such small incremental increases are not really an exciting upgrade.

 

Over-hyping Skylake? ☑

DirectX 12 is probably the best thing, in my opinion, as a gamer. But even that isn't worth waiting a year for.

 

With this said, I'm gonna build a new system with a 5930K then, thank you for helping me clear my mind

 

But we'll need to wait for a good while till games make use of it which is ok since it will give Winblows 10 a year or so at least to stabilize from the joke of an OS it is now even a few builds before we hit RTM

 

I edited my post.

 

Same here, a possible 70% performance boost according to some sites, so even if it's a fraction of that it'll be worth looking forward to!

Pleasure Enjoy that beast of a machine.

 

Thank you sir, rep added for editing your post

 

70% performance boost? On what, a Pentium G3258 at stock in BF4?

 

Not really sure what on specifically, but if you look around for the performance boost that DX12 allows, sites seem to claim 50%+ performance improvements for CPUs in games and similar.

 

Take that with massive grains of salt. I remember how "Mantle" was supposedly a "60% boost" and it was in fact using an APU to play BF4 and going from like 18fps to 30fps.

Listen, lemme make something clear. Whatever boosts you saw with Mantle is what you will see with DX12/Vulcan. The only difference is that DX12/Vulkan will have extra technologies in them, but otherwise they'll basically do what Mantle did, in terms of CPU performance benefits.

 

What's interesting after Skylake anyway?

 

Sounds like a big fib, or at least way inaccurate. No way. Maybe they meant 7.0% and forgot the decimal point. I'd also have to wonder "compared to what" LOL. Maybe an i7 Skylake does offer a 70% increase over an Intel "U" dual core netbook ultrabook processor.

 

Yeah right...more like 70% increase over a Pentium M. Who are they trying to fool here? Remember when ivy bridge was supposed to give 5-10% IPC improvement at HALF the power consumption of sandy bridge? That sure didn't happen...

 

I think the 70% mark indicates IGP performance increase.
That, or it could be 60% performance increase for IGP and 10% increase for the CPU (because I think the Broadwell top end IGP was tested and it offered about 2x performance of Haswell IGP - or something in that ballpark).

 

If Broadwell's iGPU is Iris Pro, then it needs to be tested vs an Iris Pro (mobile chip). Can't say "HD4600 vs Iris Pro 6200" and then claim benefits if Iris Pro 5200 exists to be tested, etc.

 

Agreed. The performance numbers I provided were based on what seemed to be desktop Broadwell. How mobile behaves is a different matter though given other constraints that apply to this form factor. Plus I doubt that skylake would necessarily offer significant changes in igp from broadwell. The changes would likely be on the cpu side for Skylake while the igp seems likely to carry over from broadwell (given that it already sports most/all enhancements)

 

Actually as far as early benchmarks went, Skylake desktop "K" chips had no iGPU. I distinctly remember seeing an early skylake Cinebench benchmark and they left the iGPU slot blank and claimed there was no iGPU in the CPU. Of course until Skylake launches this remains to be seen... we know the mobile chips WILL have an iGPU.

 

An raised TDP in Skylake desktop "K" chip indicates iGPU .

 

I hope they do not have an iGPU. I wish they had a fully unlocked mobile Extreme version without any iGPU also. This is just a waste of silicon on an enthusiast machine, especially so if it is a desktop CPU. They could add a couple of cores to make a hexacore mobile Extreme or something actually more useful in the eyes of an enthusiast.

 

I think unfortunately Intel focus more and more on iGPU in both desktop and mobile platform. They can do this as long as they don't have competition in terms of processing power .

 

I personally love the idea of iGPU in every laptop. First of all, battery life. Then, if crappily soldered dGPU on your mobo dies, you still have a working machine. Want to sell you MXM card? No problem - if your iGPU is not completely disabled that is (hello, P75XZM owners). Though I understand all the hate, iGPUs make my life easier.

 

An iGPU is OK as long as it doesn't get in the way or result in a performance compromise by virtue of its existence. On my Alienware SLI beasts I forget that it exists, almost never use it, so it's not an issue just as it is not for 17" Alienware owner's with 120HZ displays. Since it is disabled 24/7 unless I manually enable it and there is never any Optimus garbage causing issues, it doesn't bother me. However, if I could trade the silicon real estate for a couple of extra CPU cores, I'd totally be down with having no iGPU. The primary benefit to having it is being able to use it for re-flashing soft-bricked MXM modules. Once I get my hot air station and chip programming tools and learn how to use them, I won't need it for that any more either.

What would be nice is having a really dirt cheap generic Intel HD Graphics MXM module with about 1GB of vRAM that you could buy for like $50 to $75 and slap into an empty MXM/PCI-e slot if you needed to sell or replace a high performance GPU. Then you could use any dual MXM laptop in the same manner as I am able to with my Alienware beasts for re-flashing soft-bricked GPUs.

 

Fox, you cannot overlook the benefit of the iGPU as an asset. When I had a bad flash on my 7970M, running the iGPU allowed me to flash a proper vbios back to my 7970M which would have otherwise been impossible. It also allows your computer to run on limp mode should your MXM card fail. It also extends battery life, which greatly helps in the laptop specific applications. iGPU is a wonderful tool, but it should remain within the control of the end user whether or not it is enabled and should never be implemented in a way that would hinder dGPU performance. Enduro/Optimus are total garbage, and I long for the return of a MUX switch system for iGPU/dGPU.

Also, you do not need a dual MXM system to un-brick GPUs. I was able to do so on my M17x R4 after a bad flash simply by running SG mode, enabling both iGPU and PEG long enough for me to get into DOS and fix the bad flash.

 

Wow, that is great news. First time I ever heard that this is possible. I've fixed soft-bricked 7970M GPUs for a couple of M17xR4 owners that we all know here. Apparently, they did not know either, otherwise they would not have wasted money shipping them to me. Is this a recent discovery or something that I missed conversation about?

What is the normal setting for Optimus on the M17xR4? "Auto" instead of SG?

 

Normal setting is SG. I made the mistake of flashing an older 7970M bios on a rev 022 7970M, so naturally it wouldn't boot into windows anymore under any sort of hardware (or software) graphics acceleration when running PEG only. All I had to do was find a 022 BIOS, set to SG in bios, boot up the USB tool and reflash from there. Worked fine for me and I was able to boot back under PEG with full hardware accelerated graphics.

 

Look at this clown he seems to have a deal with the future angels....
http://www.overclock.net/t/1550982/x99-platform#post_24074975

 

Maybe not the best thread but; seems like Cannonlake is getting pushed to 2017. Next year we´ll get 14 nm Kaby Lake.

Sweclockers
Kitguru

 

Yeah... that seems like a poorly put together joke that some people fell for and didn't bother to verify.