Skylake: i5-6400 vs i7-6700T specs

[b3nbranch]

OK, I've been reading the news about the newly released Skylake specs.
If the i7-6700T has more oomph (threads, clocks, L3 Cache) than the i5-6400,
but is 35W vs 65W TDP,

Core i5-6400 14nm 4/4 2.7 GHz 3.3 GHz 6 MB DDR4 2133 MHz 65W LGA 1151 No
Core i7-6700T 14nm 4/8 2.8 GHz 3.6 GHz 8 MB DDR4 2133 MHz 35W LGA 1151 No

How has Intel managed this?

[CA_Steve]

source?

[b3nbranch]

http://wccftech.com/intels-6th-generati ... s-spotted/

http://www.cpu-world.com/CPUs/Core_i7/I ... 6700T.html

[CA_Steve]

Could be a couple of things (with the caveat that we really don't know if these are final specs)
- Intel throws CPUs into a few large TDP buckets. The two parts you mention are at the top end of the lowest TDP bucket and the lowest speed part in the middle TDP bucket. So, there may not be as huge of a difference in power use as it looks. The i5 might come in at 40W <shrugs>
- We don't know the algorithm used for Turbo. The i7 T part might have a more aggressive return to lowest speed function.
- We don't know the speed of the iGP or the number of EUs. The T part might be nerfed.

[b3nbranch]

All good points. IMHO it's likely that the T part will de-turbo if more than one core is busy.
Might also be binned for lower voltages.

Even if nerfed, the iGPU will probably run rings around the Nvidia nforce 730i iGPU in my
current system ... will be interesting to see final specs *and prices*.

[CA_Steve]

I don't think Intel has binned for lower voltage consumer parts in many years. The last few generations of parts have all followed the same V-F curve whether they were T, S, or K parts.

[alecmg]

viewtopic.php?f=28&t=68475
A test showed lower binned parts on desktop act just like their big brothers with a tiny bit less agressive turbo policy
(given adequate cooling obviously)
So no magic is happening