Say what you want about dGPUs making sense (or not) -- in Ultrabooks, this move limits the choices of consumers and prevents OEMs from designing certain products (ultrabooks) that include components from some of Intel's competitors. That is the cost we all bear, I wonder what the benefit it.
I wonder how much power they save limiting this option?
Say what you want about dGPUs making sense (or not) -- in Ultrabooks, this move limits the choices of consumers and prevents OEMs from designing certain products (ultrabooks) that include components from some of Intel's competitors. That is the cost we all bear, I wonder what the benefit it.
Until someone can figure that out, we have no way of knowing Intel's motivations. Although certainly it could be a case of two birds one stone![]()
When talking about ULVs with discrete GPUs one can't forget Razer Edge Pro. Guess the next iteration of that will be IGP-only if it uses Haswell.
Nobody who bought an ultrabook has a clue about PC hardware let's be honest.
Intel are just doing what they've always done. Nvidia were used as a toy to keep AMD in check for years, but they serve no purpose now and they'll be shut out of more markets as time goes by. The more you compete with Intel, the more they shut you out or look for alternatives or help your competition.
Only a handful, but they are out there.
So the market did decide? It sounds like its a niche that couldnt substain itself. It would have to be fitted with a 640M or higher today to matter. And the 640M is 35W TDP. Thats most likely more than the entire rest of the motherboard combined.
I really don't see this "change" being much of a change at all, because affected systems (tablets, ultra slim products) really don't need dGPU.
Is it Intel's place to decide what we really need? They seem to have decided that ultraslim products need dGPU level performance, as they have crammed GT3 into a ULV chip- but they don't want anyone but Intel offering it.
Is it Intel's place to decide what we really need? They seem to have decided that ultraslim products need dGPU level performance, as they have crammed GT3 into a ULV chip- but they don't want anyone but Intel offering it.
Wonder how this fits with the FTC settlement:
"maintain a key interface, known as the PCI Express Bus, for at least six years in a way that will not limit the performance of graphics processing chips. These assurances will provide incentives to manufacturers of complementary, and potentially competitive, products to Intels CPUs to continue to innovate; and"
6 years from 2010 = 2016
http://www.ftc.gov/opa/2010/08/intel.shtm
How many ultrabooks today are sold with discrete GFX?
Lenovo will be introducing an AMD Kabini 15" S415 model laptop with discrete Radeon 8000M graphics that boosts performance and will probably be priced much lower than Intel Ultrabooks since chips like the i7-4500u cost almost $400!
Having ability to use discrete GPUs is a good choice and I am glad AMD is providing choices to customers.
![]()
Wonder how this fits with the FTC settlement:
"maintain a key interface, known as the PCI Express Bus, for at least six years in a way that will not limit the performance of graphics processing chips. These assurances will provide incentives to manufacturers of complementary, and potentially competitive, products to Intels CPUs to continue to innovate; and"
6 years from 2010 = 2016
http://www.ftc.gov/opa/2010/08/intel.shtm
Do regular HW mobile chips support s0ix sleep states? If not then that's a bummer, the biggest improvement will only be seen in battery life on 1 chip ULV skus.The rest won't see much of an improvement, just like on the desktop. Seems like if you want dGPU, 37W TDP HW SKU is the lowest. Did anyone test HW extreme on mobile? What's that SKU name and clocks? I wonder what kind of an improvement we can expect over 3940XM.
UPDATE:
http://ark.intel.com/products/75133...essor-Extreme-Edition-8M-Cache-up-to-3_90-GHz
The exact same clocks, 3/3.9 so it should be around 10% faster. For 1100$ they really should have included Crystalwell though, even if only for CPU performance. I'd be not surprised if i7-4950HQ Processor actually won some CPU benchmarks probably for half the price.
Its clocks are 2.4/3.6 so it could potentially win quite a few ST benchmarks. It's got 10W lower TDP to boot.
Actually Haswell's S0ix is not supported in Windows 8 (the support will arrive in Win 8.1). So it is reasonable to expect similar battery life gain for non-ult parts (at least in light workloads).
Also S0ix makes little sense for non-SoC devices, so non-ult/ulx parts wont support it. S0ix are system states (not cpu states), and cpu integrated power management controller can not control external devices sitting on pcie bus (i guess one more reason Haswell ult does not support pcie).
Also you may check MS specifications for connected standby:
http://msdn.microsoft.com/en-us/library/windows/hardware/jj248729.aspx
And here is Atom z2760 power management implementation (should be similar in Haswell SoC):
https://www-ssl.intel.com/content/d...ments/product-briefs/atom-z2760-datasheet.pdf
Windows 8 also plays a significant role in all of this as the OS supports coalescing of tasks in software to ensure that it isnt working against Intels Power Optimizer in hardware.
So it looks like those states are already improving battery life in W8 but not in W7 which I still use.No Connected Standby for Now
Although you should be able to realize some of the benefits from S0ix with the first Haswell Ultrabooks, Connected Standby (periodic content refresh while in a sleep state) requires OS support. For Haswell Ultrabooks, that means waiting for Windows 8.1.
Even on the S0ix side, its not clear to me whether all of the devices in the system capable of going into D3 (their lowest power state) while active will actually do so until Windows 8.1. I get the distinct impression that Haswell Ultrabooks will see a tangible increase in battery life with Windows 8.1.
To my knowledge, that system is using the quad core mobile Haswell which is unaffected by this change. A Haswell successor for the Razer Edge is already in production and uses a GTX 775M...
Anand says otherwise in his HW ULT article.
http://www.anandtech.com/show/7047/the-haswell-ultrabook-review-core-i74500u-tested/3
He says it IS supported but we should still see tangible power consumption improvement when W8.1 comes.
Relevant parts
![]()
That's IDF slide.
So it looks like those states are already improving battery life in W8 but not in W7 which I still use.
Well you're only getting this processor in an Ultrabook and if you're buying a Haswell UltraBook, it comes with Windows 8, so kind of a non issue there.
Actually Haswell's S0ix is not supported in Windows 8 (the support will arrive in Win 8.1). So it is reasonable to expect similar battery life gain for non-ult parts (at least in light workloads).
You understand that this is being aimed at devices with sub 15W TDP? Who the heck would want to add a 35W discrete graphics chip in a tablet? A tablet doesn't have the space for a large laptop battery so it makes absolute ZERO sense to add discrete graphics to such a device - It's common sense. Aside from this, intel has to make trade-offs to improve efficiency. This is all part of that, they can't be held back by legacy - they have to change the architecture and make trade-offs to improve their efficiency.
If you want discrete graphics get a full size laptop. You don't purchase a device that is the size of an ipad for discrete graphics performance.
